id stringlengths 16 29 | text stringlengths 86 3.49k | source stringlengths 14 112 |
|---|---|---|
arxiv_dataset-172002209.14285 | Polarized gluon distribution in the proton from holographic light-front
QCD
hep-ph hep-th nucl-th
We obtain the gluon parton distribution functions (PDFs) in the proton within
the extended light-front holographic QCD framework, where the proton couples
with the spin-two Pomeron in Anti-de Sitter space, together with constraints
imposed by the Veneziano model. The gluon helicity asymmetry, after satisfying
the perturbative QCD constraints at small and large longitudinal momentum
regions, agrees with existing experimental measurements. The polarized gluon
distribution is consistent with global analyses. We predict the gluon helicity
contribution to the proton spin, $\Delta G=0.221^{+0.056}_{-0.044}$, close to
the recent analysis with updated data sets and PHENIX measurement and the
lattice QCD simulations. We subsequently present the unpolarized and polarized
gluon generalized parton distributions in the proton.
| arxiv topic:hep-ph hep-th nucl-th |
arxiv_dataset-172012209.14385 | Feature Decoupling in Self-supervised Representation Learning for Open
Set Recognition
cs.CV cs.AI cs.CR cs.LG
Assuming unknown classes could be present during classification, the open set
recognition (OSR) task aims to classify an instance into a known class or
reject it as unknown. In this paper, we use a two-stage training strategy for
the OSR problems. In the first stage, we introduce a self-supervised feature
decoupling method that finds the content features of the input samples from the
known classes. Specifically, our feature decoupling approach learns a
representation that can be split into content features and transformation
features. In the second stage, we fine-tune the content features with the class
labels. The fine-tuned content features are then used for the OSR problems.
Moreover, we consider an unsupervised OSR scenario, where we cluster the
content features learned from the first stage. To measure representation
quality, we introduce intra-inter ratio (IIR). Our experimental results
indicate that our proposed self-supervised approach outperforms others in image
and malware OSR problems. Also, our analyses indicate that IIR is correlated
with OSR performance.
| arxiv topic:cs.CV cs.AI cs.CR cs.LG |
arxiv_dataset-172022209.14485 | A Semi-Analytical Line Transfer (SALT) Model III: Galactic Inflows
astro-ph.GA
We present calculations of ultraviolet spectra resulting from the scattering
of photons by gas in-falling onto an isotropically emitting source of
radiation. The model is based on an adaptation of the semi-analytical line
transfer (SALT) code of Scarlata & Panagia (2015), and designed to interpret
the inverse P-Cygni profiles observed in the spectra of partially ionized
galactic inflows. In addition to presenting the model, we explore the parameter
space of the inflowing SALT model and recreate various physically motivated
scenarios including spherical inflows, inflows with covering fractions less
than unity, and galactic fountains (i.e., galactic systems with both an
inflowing and outflowing component). The resulting spectra from inflowing gas
show spectral features that could be misinterpreted as ISM features in low
resolution spectroscopy ($\sigma \approx 120$ $\rm{km }$ $\rm{s}^{-1}$),
suggesting that the total number of galactic systems with inflows is
undercounted. Our models suggest that observations at medium resolution ($R =
6000$ or $\sigma \approx 50$ $\rm{km }$ $\rm{s}^{-1}$) that can be obtained
with 8m-class telescopes will be able to resolve the characteristic inverse P
Cygni profiles necessary to identify inflows.
| arxiv topic:astro-ph.GA |
arxiv_dataset-172032209.14585 | Ultra-massive spacetimes
gr-qc hep-th
A positive cosmological constant $\Lambda >0$ sets an upper limit for the
area of marginally future-trapped surfaces enclosing a black hole (BH). Does
this mean that the mass of the BH cannot increase beyond the corresponding
limit? I analyze some simple spherically symmetric models where regions within
a dynamical horizon keep gaining mass-energy so that eventually the $\Lambda$
limit is surpassed. This shows that the black hole proper transmutes into a
collapsing universe, and no observers will ever reach infinity, which
dematerializes together with the event horizon and the `cosmological horizon'.
The region containing the dynamical horizon cannot be causally influenced by
the vast majority of the spacetime, its past being just a finite portion of the
total, spatially infinite, spacetime. Thereby, a new type of horizon arises,
but now relative to past null infinity: the boundary of the past of all
marginally trapped spheres, which contains in particular one with the maximum
area $4\pi/\Lambda$. The singularity is universal and extends mostly outside
the collapsing matter. The resulting spacetimes models turn out to be
inextendible and globally hyperbolic. It is remarkable that they cannot exist
if $\Lambda$ vanishes. Given the accepted value of $\Lambda$ deduced from
cosmological observations, such ultra-massive objects will need to contain a
substantial portion of the total {\it present} mass of the {\it observable}
Universe.
| arxiv topic:gr-qc hep-th |
arxiv_dataset-172042209.14685 | Self-stabilizing Total-order Broadcast
cs.DC
The problem of total-order (uniform reliable) broadcast is fundamental in
fault-tolerant distributed computing since it abstracts a broad set of problems
requiring processes to uniformly deliver messages in the same order in which
they were sent. Existing solutions (that tolerate process failures) reduce the
total-order broadcast problem to the one of multivalued consensus.
Our study aims at the design of an even more reliable solution. We do so
through the lenses of self-stabilization-a very strong notion of fault
tolerance. In addition to node and communication failures, self-stabilizing
algorithms can recover after the occurrence of arbitrary transient faults;
these faults represent any violation of the assumptions according to which the
system was designed to operate (as long as the algorithm code stays intact).
This work proposes the first (to the best of our knowledge) self-stabilizing
algorithm for total-order (uniform reliable) broadcast for asynchronous
message-passing systems prone to process failures and transient faults. As we
show, the proposed solution facilitates the elegant construction of
self-stabilizing state-machine replication using bounded memory.
| arxiv topic:cs.DC |
arxiv_dataset-172052209.14785 | EMF-Aware MU-MIMO Beamforming in RIS-Aided Cellular Networks
cs.IT eess.SP math.IT
Reconfigurable Intelligent Surfaces (RISs) are one of the key emerging 6th
Generation (6G) technologies that are expected to improve the link budgets
between transmitters and receivers by adding artificial propagation paths. In
such re-configured propagation environment, Downlink (DL) Multi-User
Multi-Input Multi-Output (MU-MIMO) brings capacity improvement to cellular
networks. It benefits from the spatial dimension offered by MIMO systems to
enable simultaneous transmission of independent data streams to multiple users
on the same radio resources by applying appropriate Beamforming (BF) schemes.
However, in some cases, serving the same subset of users for a long period of
time may cause some undesired regions where the average Electromagnetic Field
Exposure (EMFE) exceeds the regulatory limits. To address this challenge, we
propose in this paper a novel Electromagnetic Field (EMF) aware MU-MIMO BF
scheme that aims to optimize the overall capacity under EMF constraints in
RIS-aided cellular networks.
| arxiv topic:cs.IT eess.SP math.IT |
arxiv_dataset-172062209.14885 | Majorana corner modes and flat-band Majorana edge modes in
superconductor/topological-insulator/superconductor junctions
cond-mat.mes-hall
Recently, superconductors with higher-order topology have stimulated
extensive attention and research interest. Higher-order topological
superconductors exhibit unconventional bulk-boundary correspondence, thus allow
exotic lower-dimensional boundary modes, such as Majorana corner and hinge
modes. However, higher-order topological superconductivity has yet to be found
in naturally occurring materials. In this work, we investigate higher-order
topology in a two-dimensional Josephson junction comprised of two $s$-wave
superconductors separated by a topological insulator thin film. We found that
zero-energy Majorana corner modes, a boundary fingerprint of higher-order
topological superconductivity, can be achieved by applying magnetic field. When
an in-plane Zeeman field is applied to the system, two corner states appear in
the superconducting junction. Furthermore, we also discover a two dimensional
nodal superconducting phase which supports flat-band Majorana edge modes
connecting the bulk nodes. Importantly, we demonstrate that zero-energy
Majorana corner modes are stable when increasing the thickness of topological
insulator thin film.
| arxiv topic:cond-mat.mes-hall |
arxiv_dataset-172072209.14985 | Nonprobabilistic typicality with application to quantum mechanics
quant-ph
In this paper two hypotheses are developed. The first hypothesis is the
existence of random phenomena/experiments in which the events cannot generally
be assigned a definite probability but that nevertheless admit a class of
nearly certain events. These experiments are referred to as
\textit{typicalistic} (instead of probabilistic) experiments. As probabilistic
experiments are represented by probability spaces, typicalistic experiments can
be represented by \textit{typicality spaces}, where a typicality space is
basically a probability space in which the probability measure has been
replaced by a much less structured typicality measure $T$. The condition $T(A)
\approx 1$ defines the typical sets, and a typicality space is related to a
typicalistic experiment by associating the typical sets of the former with the
nearly certain events of the latter. Some elements of a theory of typicality,
including the definition of typicality spaces, are developed in the first part
of the paper. The second hypothesis is that the evolution of a quantum particle
(or of a system of quantum particles) can be considered a typicalistic
phenomenon, so that it can be represented by the combination of typicality
theory and quantum mechanics. The result is a novel formulation of quantum
mechanics that does not present the measurement problem and that could be a
viable alternative to Bohmian mechanics. This subject is developed in the
second part of the paper.
| arxiv topic:quant-ph |
arxiv_dataset-172082209.15085 | Cardiotocography Signal Abnormality Detection based on Deep Unsupervised
Models
eess.SP
Cardiotocography (CTG) is a key element when it comes to monitoring fetal
well-being. Obstetricians use it to observe the fetal heart rate (FHR) and the
uterine contraction (UC). The goal is to determine how the fetus reacts to the
contraction and whether it is receiving adequate oxygen. If a problem occurs,
the physician can then respond with an intervention. Unfortunately, the
interpretation of CTGs is highly subjective and there is a low inter- and
intra-observer agreement rate among practitioners. This can lead to unnecessary
medical intervention that represents a risk for both the mother and the fetus.
Recently, computer-assisted diagnosis techniques, especially based on
artificial intelligence models (mostly supervised), have been proposed in the
literature. But, many of these models lack generalization to unseen/test data
samples due to overfitting. Moreover, the unsupervised models were applied to a
very small portion of the CTG samples where the normal and abnormal classes are
highly separable. In this work, deep unsupervised learning approaches, trained
in a semi-supervised manner, are proposed for anomaly detection in CTG signals.
The GANomaly framework, modified to capture the underlying distribution of data
samples, is used as our main model and is applied to the CTU-UHB dataset.
Unlike the recent studies, all the CTG data samples, without any specific
preferences, are used in our work. The experimental results show that our
modified GANomaly model outperforms state-of-the-arts. This study admit the
superiority of the deep unsupervised models over the supervised ones in CTG
abnormality detection.
| arxiv topic:eess.SP |
arxiv_dataset-172092209.15185 | Dephasing effect promotes the appearance of quantized Hall plateaus
cond-mat.mes-hall
The quantum Hall effect (QHE) is a topologically protected phenomenon which
has been observed in various systems. In experiments, the size of Hall bar
device to realize the QHE is generally much larger than the phase coherence
length, in which the quantum coherence of electrons is destroyed. Here, we
theoretically study the influence of dephasing effect on the quantized Hall
(QH) plateaus. We find that the QH plateau disappears in perfectly quantum
coherent systems if the coupling between leads and central region is imperfect.
The Hall resistance is very large and strongly oscillates instead of presenting
the QH plateau in this case. However, by introducing the dephasing, the Hall
resistance decreases and the QH plateau appears gradually. Similar results can
also be observed for the quantum anomalous Hall effect. Our results propose
that dephasing effect promotes the appearance of QH plateaus, which opens a new
topic of the dephasing effect on topological systems.
| arxiv topic:cond-mat.mes-hall |
arxiv_dataset-172102209.15285 | QUAK: A Synthetic Quality Estimation Dataset for Korean-English Neural
Machine Translation
cs.CL
With the recent advance in neural machine translation demonstrating its
importance, research on quality estimation (QE) has been steadily progressing.
QE aims to automatically predict the quality of machine translation (MT) output
without reference sentences. Despite its high utility in the real world, there
remain several limitations concerning manual QE data creation: inevitably
incurred non-trivial costs due to the need for translation experts, and issues
with data scaling and language expansion. To tackle these limitations, we
present QUAK, a Korean-English synthetic QE dataset generated in a fully
automatic manner. This consists of three sub-QUAK datasets QUAK-M, QUAK-P, and
QUAK-H, produced through three strategies that are relatively free from
language constraints. Since each strategy requires no human effort, which
facilitates scalability, we scale our data up to 1.58M for QUAK-P, H and 6.58M
for QUAK-M. As an experiment, we quantitatively analyze word-level QE results
in various ways while performing statistical analysis. Moreover, we show that
datasets scaled in an efficient way also contribute to performance improvements
by observing meaningful performance gains in QUAK-M, P when adding data up to
1.58M.
| arxiv topic:cs.CL |
arxiv_dataset-172112209.15385 | On the number of solutions to the planar dual Minkowski problem
math.AP math.DG
The dual Minkowski problem in the two-dimensional plane is studied in this
paper. By combining the theoretical analysis and numerical estimation of an
integral with parameters, we find the number of solutions to this problem for
the constant dual curvature case when $0<q\leq4$. An improved nonuniqueness
result when $q>4$ is also obtained. As an application, a result on the
uniqueness and nonuniqueness of solutions to the $L_p$-Alexandrov problem is
obtained for $p<0$.
| arxiv topic:math.AP math.DG |
arxiv_dataset-172122209.15485 | A New Group of Two-Dimensional Non-van der Waals Materials with Ultra
Low Exfoliation Energies
cond-mat.mtrl-sci physics.comp-ph
The exfoliation energy - quantifying the energy required to extract a
two-dimensional (2D) sheet from the surface of a bulk material - is a key
parameter determining the synthesizability of 2D compounds. Here, using ab
initio calculations, we present a new group of non-van der Waals 2D materials
derived from non-layered crystals which exhibit ultra low exfoliation energies.
In particular for sulfides, surface relaxations are essential to correctly
describe the associated energy gain needed to obtain reliable results. Taking
into account long-range dispersive interactions has only a minor effect on the
energetics and ultimately proves that the exfoliation energies are close to the
ones of traditional van der Waals bound 2D compounds. The candidates with the
lowest energies, 2D SbTlO$_3$ and MnNaCl$_3$, exhibit appealing electronic,
potential topological, and magnetic features as evident from the calculated
band structures making these systems an attractive platform for fundamental and
applied nanoscience.
| arxiv topic:cond-mat.mtrl-sci physics.comp-ph |
arxiv_dataset-172132209.15585 | Cloud Classification with Unsupervised Deep Learning
physics.ao-ph cs.LG
We present a framework for cloud characterization that leverages modern
unsupervised deep learning technologies. While previous neural network-based
cloud classification models have used supervised learning methods, unsupervised
learning allows us to avoid restricting the model to artificial categories
based on historical cloud classification schemes and enables the discovery of
novel, more detailed classifications. Our framework learns cloud features
directly from radiance data produced by NASA's Moderate Resolution Imaging
Spectroradiometer (MODIS) satellite instrument, deriving cloud characteristics
from millions of images without relying on pre-defined cloud types during the
training process. We present preliminary results showing that our method
extracts physically relevant information from radiance data and produces
meaningful cloud classes.
| arxiv topic:physics.ao-ph cs.LG |
arxiv_dataset-172142210.00045 | Calibrating Sequence likelihood Improves Conditional Language Generation
cs.CL
Conditional language models are predominantly trained with maximum likelihood
estimation (MLE), giving probability mass to sparsely observed target
sequences. While MLE trained models assign high probability to plausible
sequences given the context, the model probabilities often do not accurately
rank-order generated sequences by quality. This has been empirically observed
in beam search decoding as output quality degrading with large beam sizes, and
decoding strategies benefiting from heuristics such as length normalization and
repetition-blocking. In this work, we introduce sequence likelihood calibration
(SLiC) where the likelihood of model generated sequences are calibrated to
better align with reference sequences in the model's latent space. With SLiC,
decoding heuristics become unnecessary and decoding candidates' quality
significantly improves regardless of the decoding method. Furthermore, SLiC
shows no sign of diminishing returns with model scale, and presents alternative
ways to improve quality with limited training and inference budgets. With SLiC,
we exceed or match SOTA results on a wide range of generation tasks spanning
abstractive summarization, question generation, abstractive question answering
and data-to-text generation, even with modest-sized models.
| arxiv topic:cs.CL |
arxiv_dataset-172152210.00145 | Coalitional Game-Theoretical Approach to Coinvestment with Application
to Edge Computing
cs.GT cs.NI
We propose in this paper a coinvestment plan between several stakeholders of
different types, namely a physical network owner, operating network nodes, e.g.
a network operator or a tower company, and a set of service providers willing
to use these resources to provide services as video streaming, augmented
reality, autonomous driving assistance, etc. One such scenario is that of
deployment of Edge Computing resources.
Indeed, although the latter technology is ready, the high Capital Expenditure
(CAPEX) cost of such resources is the barrier to its deployment. For this
reason, a solid economical framework to guide the investment and the returns of
the stakeholders is key to solve this issue. We formalize the coinvestment
framework using coalitional game theory. We provide a solution to calculate how
to divide the profits and costs among the stakeholders, taking into account
their characteristics: traffic load, revenues, utility function. We prove that
it is always possible to form the grand coalition composed of all the
stakeholders, by showing that our game is convex. We derive the payoff of the
stakeholders using the Shapley value concept, and elaborate on some properties
of our game. We show our solution in simulation.
| arxiv topic:cs.GT cs.NI |
arxiv_dataset-172162210.00245 | Uniqueness for 2-Intersecting Families of Permutations and Perfect
Matchings
math.CO
We give a characterization of the largest $2$-intersecting families of
permutations of $\{1,2,\ldots,n\}$ and of perfect matchings of the complete
graph $K_{2n}$ for all $n \geq 2$.
| arxiv topic:math.CO |
arxiv_dataset-172172210.00345 | Identifying Selections Operating on HIV-1 Reverse Transcriptase via
Uniform Manifold Approximation and Projection
q-bio.GN cs.CE cs.LG
We analyze 14,651 HIV1 reverse transcriptase (HIV RT) sequences from the
Stanford HIV Drug Resistance Database labeled with treatment regimen in order
to study the evolution this enzyme under drug selection in the clinic. Our goal
is to identify distinct sectors of HIV RT's sequence space that are undergoing
evolution as a way to identify individual selections and/or evolutionary
solutions. We utilize Uniform Manifold Approximation and Projection (UMAP), a
graph-based dimensionality reduction technique uniquely suited for the
detection of non-linear dependencies and visualize the results using an
unsupervised clustering algorithm based on density analysis. Our analysis
produced 21 distinct clusters of sequences. Supporting the biological
significance of these clusters, they tend to represent phylogenetically related
sequences with strong correspondence to distinct treatment regimens. Thus, this
method for visualization of areas of HIV RT undergoing evolution can help infer
information about selective pressures, although it is correlative. The mutation
signatures associated with each cluster may represent the higher-order
epistatic context facilitating these evolutionary pathways, information that is
generally not accessible by other types of mutational co-dependence analyses.
| arxiv topic:q-bio.GN cs.CE cs.LG |
arxiv_dataset-172182210.00445 | ManiCLIP: Multi-Attribute Face Manipulation from Text
cs.CV
In this paper we present a novel multi-attribute face manipulation method
based on textual descriptions. Previous text-based image editing methods either
require test-time optimization for each individual image or are restricted to
single attribute editing. Extending these methods to multi-attribute face image
editing scenarios will introduce undesired excessive attribute change, e.g.,
text-relevant attributes are overly manipulated and text-irrelevant attributes
are also changed. In order to address these challenges and achieve natural
editing over multiple face attributes, we propose a new decoupling training
scheme where we use group sampling to get text segments from same attribute
categories, instead of whole complex sentences. Further, to preserve other
existing face attributes, we encourage the model to edit the latent code of
each attribute separately via an entropy constraint. During the inference
phase, our model is able to edit new face images without any test-time
optimization, even from complex textual prompts. We show extensive experiments
and analysis to demonstrate the efficacy of our method, which generates natural
manipulated faces with minimal text-irrelevant attribute editing. Code and
pre-trained model are available at https://github.com/hwang1996/ManiCLIP.
| arxiv topic:cs.CV |
arxiv_dataset-172192210.00545 | Seeing Through the Noisy Dark: Towards Real-world Low-Light Image
Enhancement and Denoising
cs.CV
Low-light image enhancement (LLIE) aims at improving the illumination and
visibility of dark images with lighting noise. To handle the real-world
low-light images often with heavy and complex noise, some efforts have been
made for joint LLIE and denoising, which however only achieve inferior
restoration performance. We attribute it to two challenges: 1) in real-world
low-light images, noise is somewhat covered by low-lighting and the left noise
after denoising would be inevitably amplified during enhancement; 2) conversion
of raw data to sRGB would cause information loss and also more noise, and hence
prior LLIE methods trained on raw data are unsuitable for more common sRGB
images. In this work, we propose a novel Low-light Enhancement & Denoising
Network for real-world low-light images (RLED-Net) in the sRGB color space. In
RLED-Net, we apply a plug-and-play differentiable Latent Subspace
Reconstruction Block (LSRB) to embed the real-world images into low-rank
subspaces to suppress the noise and rectify the errors, such that the impact of
noise during enhancement can be effectively shrunk. We then present an
efficient Crossed-channel & Shift-window Transformer (CST) layer with two
branches to calculate the window and channel attentions to resist the
degradation (e.g., speckle noise and blur) caused by the noise in input images.
Based on the CST layers, we further present a U-structure network CSTNet as
backbone for deep feature recovery, and construct a feature refine block to
refine the final features. Extensive experiments on both real noisy images and
public image databases well verify the effectiveness of the proposed RLED-Net
for RLLIE and denoising simultaneously.
| arxiv topic:cs.CV |
arxiv_dataset-172202210.00645 | Economic-Driven Adaptive Traffic Signal Control
cs.MA
With the emerging connected-vehicle technologies and smart roads, the need
for intelligent adaptive traffic signal controls is more than ever before. This
paper proposes a novel Economic-driven Adaptive Traffic Signal Control (eATSC)
model with a hyper control variable - interest rate defined in economics for
traffic signal control at signalized intersections. The eATSC uses a continuous
compounding function that captures both the total number of vehicles and the
accumulated waiting time of each vehicle to compute penalties for different
directions. The computed penalties grow with waiting time and is used for
signal control decisions. Each intersection is assigned two intelligent agents
adjusting interest rate and signal length for different directions according to
the traffic patterns, respectively. The problem is formulated as a Markov
Decision Process (MDP) problem to reduce congestions, and a two-agent Double
Dueling Deep Q Network (DDDQN) is utilized to solve the problem. Under the
optimal policy, the agents can select the optimal interest rates and signal
time to minimize the likelihood of traffic congestion. To evaluate the
superiority of our method, a VISSIM simulation model with classic four-leg
signalized intersections is developed. The results indicate that the proposed
model is adequately able to maintain healthy traffic flow at the intersection.
| arxiv topic:cs.MA |
arxiv_dataset-172212210.00745 | Electron-phonon coupling and spin fluctuations in the Ising
superconductor NbSe$_{2}$
cond-mat.supr-con cond-mat.mtrl-sci cond-mat.str-el
Ising superconductivity, observed experimentally in NbSe$_{2}$ and similar
materials, has generated tremendous interest. Recently, attention was called to
the possible role that spin fluctuations (SF) play in this phenomenon, in
addition to the dominant electron-phonon coupling (EPC); the possibility of a
predominantly-triplet state was discussed and led to a conjecture of viable
singlet-triplet Leggett oscillations. However, these hypotheses have not been
put to a quantitative test. In this paper, we report first principle
calculations of the EPC and also estimate coupling with SF, including full
momentum dependence. We find that: (1) EPC is strongly anisotropic, largely
coming from the K-K' scattering, and therefore excludes triplet symmetry even
as an excited state; (2) superconductivity is substantially weakened by SF, but
anisotropy remains as above; and, (3) we do find the possibility of a Leggett
mode, not in a singlet-triplet but in an $s_{++}$ - $s_\pm$ channel.
| arxiv topic:cond-mat.supr-con cond-mat.mtrl-sci cond-mat.str-el |
arxiv_dataset-172222210.00845 | Computational Design of Corrosion-resistant and Wear-resistant Titanium
Alloys for Orthopedic Implants
physics.med-ph
Titanium alloys are promising candidates for orthopedic implants due to their
mechanical resilience and biocompatibility. Current titanium alloys in
orthopedic implants still suffer from low wear and corrosion resistance. Here,
we present a computational method for optimizing the composition of titanium
alloys for enhanced corrosion and wear resistance without compromising on other
aspects such as phase stability, biocompatibility, and strength. We use the
cohesive energy, oxide formation energy, surface work function, and the elastic
shear modulus of pure elements as proxy descriptors to guide us towards alloys
with enhanced wear and corrosion resistance. For the best-selected candidates,
we then use the CALPHAD approach, as implemented in the Thermo-Calc software,
to calculate the phase diagram, yield strength, hardness, Pourbaix diagram, and
the Pilling-Bedworth (PB) ratio. These calculations are used to assess the
thermodynamic stability, biocompatibility, corrosion resistance, and wear
resistance of the selected alloys. Additionally, we provide insights about the
role of silicon on improving the corrosion and wear resistance of alloys.
| arxiv topic:physics.med-ph |
arxiv_dataset-172232210.00945 | Cooperative Multi-Agent Deep Reinforcement Learning for Reliable and
Energy-Efficient Mobile Access via Multi-UAV Control
cs.LG cs.MA
This paper addresses a novel multi-agent deep reinforcement learning
(MADRL)-based positioning algorithm for multiple unmanned aerial vehicles
(UAVs) collaboration (i.e., UAVs work as mobile base stations). The primary
objective of the proposed algorithm is to establish dependable mobile access
networks for cellular vehicle-to-everything (C-V2X) communication, thereby
facilitating the realization of high-quality intelligent transportation systems
(ITS). The reliable mobile access services can be achieved in following two
ways, i.e., i) energy-efficient UAV operation and ii) reliable wireless
communication services. For energy-efficient UAV operation, the reward of our
proposed MADRL algorithm contains the features for UAV energy consumption
models in order to realize efficient operations. Furthermore, for reliable
wireless communication services, the quality of service (QoS) requirements of
individual users are considered as a part of rewards and 60GHz mmWave radio is
used for mobile access. This paper considers the 60GHz mmWave access for
utilizing the benefits of i) ultra-wide-bandwidth for multi-Gbps high-speed
communications and ii) high-directional communications for spatial reuse that
is obviously good for densely deployed users. Lastly, the comprehensive and
data-intensive performance evaluation of the proposed MADRL-based algorithm for
multi-UAV positioning is conducted in this paper. The results of these
evaluations demonstrate that the proposed algorithm outperforms other existing
algorithms.
| arxiv topic:cs.LG cs.MA |
arxiv_dataset-172242210.01045 | Properties of shock waves in the quiet Sun chromosphere
astro-ph.SR
Short-lived (100s or less), sub-arcsec to a couple of arcsec size features of
enhanced brightenings in the narrowband images at the $\mathrm{H_{2V}}$ and
$\mathrm{K_{2V}}$ positions of the Ca II H and K lines in the quiet Sun are
known as bright grains.
With simultaneous observations of a quiet Sun internetwork region in the Fe I
6173 {\AA}, Ca II 8542 {AA}, and Ca II K lines acquired by the CRisp Imaging
Spectro-Polarimeter and the CHROMospheric Imaging Spectrometer instruments on
the Swedish 1-m Solar Telescope, we performed multi-line non-local
thermodynamic equilibrium inversions using the STockholm inversion Code to
infer the time-varying stratified atmosphere's physical properties such as the
temperature, line-of-sight (LOS) velocity, and microturbulence.
The Ca II K profiles of bright grains show enhancement in the
$\mathrm{K_{2V}}$ peak intensities with absence of the $\mathrm{K_{2R}}$
features. At the time of maximum enhancement in the $\mathrm{K_{2V}}$ peak
intensities, we found average enhancements in temperature at lower
chromospheric layers (at $\log\tau_{500}$ $\simeq$ $-$4.2) of about 1.1 kK with
maximum enhancement of about 4.5 kK. These temperature enhancements are
colocated with upflows, as strong as $-$6 $\mathrm{km\;s^{-1}}$, in the
direction of the LOS. The LOS velocities at upper chromospheric layers at
$\log\tau_{500}$ < $-$4.2 show consistent downflows greater than $+$8
$\mathrm{km\;s^{-1}}$. The retrieved value of microturbulence in the atmosphere
of bright grains is negligible at chromospheric layers.
The study provides observational evidence to support the interpretation that
the bright grains observed in narrowband images at the $\mathrm{H_{2V}}$ and
$\mathrm{K_{2V}}$ positions of the Ca II H and K lines are manifestations of
upward propagating acoustic shocks against a background of downflowing
atmospheres.
| arxiv topic:astro-ph.SR |
arxiv_dataset-172252210.01145 | An approximate local modular quantum energy inequality in general
quantum field theory
math-ph gr-qc hep-th math.MP
For every local quantum field theory on a static, globally hyperbolic
spacetime of arbitrary dimension, assuming the Reeh-Schlieder property, local
preparability of states, and the existence of an energy density as
operator-valued distribution, we prove an approximate quantum energy inequality
for a dense set of vector states. The quantum field theory is given by a net of
von Neumann algebras of observables, and the energy density is assumed to
fulfill polynomial energy bounds and to locally generate the time translations.
While being approximate in the sense that it is controlled by a small parameter
that depends on the respective state vector, the derived lower bound on the
expectation value of the spacetime averaged energy density has a universal
structure. In particular, the bound is directly related to the Tomita-Takesaki
modular operators associated to the local von Neumann algebras. This reveals
general, model-independent features of quantum energy inequalities for a large
class of quantum field theories on static spacetimes.
| arxiv topic:math-ph gr-qc hep-th math.MP |
arxiv_dataset-172262210.01245 | Random orthogonal additive filters: a solution to the
vanishing/exploding gradient of deep neural networks
cs.LG cs.AI cs.NA math.DS math.NA
Since the recognition in the early nineties of the vanishing/exploding (V/E)
gradient issue plaguing the training of neural networks (NNs), significant
efforts have been exerted to overcome this obstacle. However, a clear solution
to the V/E issue remained elusive so far. In this manuscript a new architecture
of NN is proposed, designed to mathematically prevent the V/E issue to occur.
The pursuit of approximate dynamical isometry, i.e. parameter configurations
where the singular values of the input-output Jacobian are tightly distributed
around 1, leads to the derivation of a NN's architecture that shares common
traits with the popular Residual Network model. Instead of skipping connections
between layers, the idea is to filter the previous activations orthogonally and
add them to the nonlinear activations of the next layer, realising a convex
combination between them. Remarkably, the impossibility for the gradient
updates to either vanish or explode is demonstrated with analytical bounds that
hold even in the infinite depth case. The effectiveness of this method is
empirically proved by means of training via backpropagation an extremely deep
multilayer perceptron of 50k layers, and an Elman NN to learn long-term
dependencies in the input of 10k time steps in the past. Compared with other
architectures specifically devised to deal with the V/E problem, e.g. LSTMs for
recurrent NNs, the proposed model is way simpler yet more effective.
Surprisingly, a single layer vanilla RNN can be enhanced to reach state of the
art performance, while converging super fast; for instance on the psMNIST task,
it is possible to get test accuracy of over 94% in the first epoch, and over
98% after just 10 epochs.
| arxiv topic:cs.LG cs.AI cs.NA math.DS math.NA |
arxiv_dataset-172272210.01345 | Lecture notes on generalized Monge-Amp\`ere equations and subvarieties
math.DG math.AG math.CV
These are the lecture notes for the Morningside Center of Mathematics
Geometry Summer School on August 15-20, 2022. These lectures sketch the results
by Yau, Demailly-Paun, the author, and Datar-Pingali about generalized
Monge-Amp\`ere equations and subvarieties and aim to use these results to study
the Hodge conjecture.
| arxiv topic:math.DG math.AG math.CV |
arxiv_dataset-172282210.01445 | A Probabilistic Model for Aircraft in Climb using Monotonic Functional
Gaussian Process Emulators
cs.CE
Ensuring vertical separation is a key means of maintaining safe separation
between aircraft in congested airspace. Aircraft trajectories are modelled in
the presence of significant epistemic uncertainty, leading to discrepancies
between observed trajectories and the predictions of deterministic models,
hampering the task of planning to ensure safe separation. In this paper a
probabilistic model is presented, for the purpose of emulating the trajectories
of aircraft in climb and bounding the uncertainty of the predicted trajectory.
A monotonic, functional representation exploits the spatio-temporal
correlations in the radar observations. Through the use of Gaussian Process
Emulators, features that parameterise the climb are mapped directly to
functional outputs, providing a fast approximation, while ensuring that the
resulting trajectory is monotonic. The model was applied as a probabilistic
digital twin for aircraft in climb and baselined against BADA, a deterministic
model widely used in industry. When applied to an unseen test dataset, the
probabilistic model was found to provide a mean prediction that was 21% more
accurate, with a 34% sharper forecast.
| arxiv topic:cs.CE |
arxiv_dataset-172292210.01545 | A Note on Entanglement Entropy for Primary Fermion Fields in JT Gravity
hep-th
In this paper we analyse and discuss 2D Jackiw-Teitelboim (JT) gravity
coupled to primary fermion fields in asymptotically anti-de Sitter (AdS)
spacetimes. We get a particular solution of the massless Dirac field outside
the extremal black hole horizon and find the solution for the dilaton in JT
gravity. Two dimensional JT gravity spacetime is conformally flat, we calculate
the two point correlators of primary fermion fields under the Weyl
transformations. The key point of this work is to present a standard technique
which is called resolvent rather than CFT methods. We redefine the fields in
terms of the conformal factor as the fermion fields, and we use the resolvent
technique to derive the renormalized entanglement entropy for massless Dirac
fields in JT gravity.
| arxiv topic:hep-th |
arxiv_dataset-172302210.01645 | Robotic Learning the Sequence of Packing Irregular Objects from Human
Demonstrations
cs.RO cs.LG
We tackle the challenge of robotic bin packing with irregular objects, such
as groceries. Given the diverse physical attributes of these objects and the
complex constraints governing their placement and manipulation, employing
preprogrammed strategies becomes unfeasible. Our approach is to learn directly
from expert demonstrations in order to extract implicit task knowledge and
strategies to ensure safe object positioning, efficient use of space, and the
generation of human-like behaviors that enhance human-robot trust.
We rely on human demonstrations to learn a Markov chain for predicting the
object packing sequence for a given set of items and then compare it with human
performance. Our experimental results show that the model outperforms human
performance by generating sequence predictions that humans classify as
human-like more frequently than human-generated sequences.
The human demonstrations were collected using our proposed VR platform,
BoxED, which is a box packaging environment for simulating real-world objects
and scenarios for fast and streamlined data collection with the purpose of
teaching robots. We collected data from 43 participants packing a total of 263
boxes with supermarket-like objects, yielding 4644 object manipulations. Our VR
platform can be easily adapted to new scenarios and objects, and is publicly
available, alongside our dataset, at https://github.com/andrejfsantos4/BoxED.
| arxiv topic:cs.RO cs.LG |
arxiv_dataset-172312210.01745 | Making Decisions under Outcome Performativity
cs.LG cs.CY stat.ML
Decision-makers often act in response to data-driven predictions, with the
goal of achieving favorable outcomes. In such settings, predictions don't
passively forecast the future; instead, predictions actively shape the
distribution of outcomes they are meant to predict. This performative
prediction setting raises new challenges for learning "optimal" decision rules.
In particular, existing solution concepts do not address the apparent tension
between the goals of forecasting outcomes accurately and steering individuals
to achieve desirable outcomes.
To contend with this concern, we introduce a new optimality concept --
performative omniprediction -- adapted from the supervised (non-performative)
learning setting. A performative omnipredictor is a single predictor that
simultaneously encodes the optimal decision rule with respect to many
possibly-competing objectives. Our main result demonstrates that efficient
performative omnipredictors exist, under a natural restriction of performative
prediction, which we call outcome performativity. On a technical level, our
results follow by carefully generalizing the notion of outcome
indistinguishability to the outcome performative setting. From an appropriate
notion of Performative OI, we recover many consequences known to hold in the
supervised setting, such as omniprediction and universal adaptability.
| arxiv topic:cs.LG cs.CY stat.ML |
arxiv_dataset-172322210.01845 | Simulation study of particle identification using cluster counting
technique for the BESIII drift chamber
hep-ex physics.ins-det
The particle identification of charged hadrons, especially for the separation
of $K$ and $\pi$, is crucial for the flavour physics study. Ionization
measurement with the cluster counting technique, which has much less
fluctuation than traditional $dE/dx$ measurement, is expected to provide better
particle identification for the BESIII experiment. Simulation studies,
including a Garfield++ based waveform analysis and a performance study of K/\pi
identification in the BESIII, offline software system have been performed. The
results show that $K/\pi$ separation power and PID efficiency would be improved
significantly in the momentum range above 1.2 GeV/c using cluster counting
technique even with conservative resolution assumption.
| arxiv topic:hep-ex physics.ins-det |
arxiv_dataset-172332210.01945 | Rigidity results for shrinking and expanding Ricci solitons
math.DG
In this paper, we prove some rigidity results for both shrinking and
expanding Ricci solitons. First, we prove that compact shrinking Ricci solitons
are Einstein if we control the maximum value of the potential function. Then,
we prove some rigidity results for non-compact gradient expanding and shrinking
Ricci solitons with pinched Ricci and scalar curvatures, assuming an asymptotic
condition on the scalar curvature at infinity.
| arxiv topic:math.DG |
arxiv_dataset-172342210.02045 | Coarse-to-Fine Point Cloud Registration with SE(3)-Equivariant
Representations
cs.CV cs.RO
Point cloud registration is a crucial problem in computer vision and
robotics. Existing methods either rely on matching local geometric features,
which are sensitive to the pose differences, or leverage global shapes, which
leads to inconsistency when facing distribution variances such as partial
overlapping. Combining the advantages of both types of methods, we adopt a
coarse-to-fine pipeline that concurrently handles both issues. We first reduce
the pose differences between input point clouds by aligning global features;
then we match the local features to further refine the inaccurate alignments
resulting from distribution variances. As global feature alignment requires the
features to preserve the poses of input point clouds and local feature matching
expects the features to be invariant to these poses, we propose an
SE(3)-equivariant feature extractor to simultaneously generate two types of
features. In this feature extractor, representations that preserve the poses
are first encoded by our novel SE(3)-equivariant network and then converted
into pose-invariant ones by a pose-detaching module. Experiments demonstrate
that our proposed method increases the recall rate by 20% compared to
state-of-the-art methods when facing both pose differences and distribution
variances.
| arxiv topic:cs.CV cs.RO |
arxiv_dataset-172352210.02145 | XMM-Newton view of the shock heating in an early merging cluster, CIZA
J1358.9$-$4750
astro-ph.GA astro-ph.HE
CIZA J1358.9-4750 is a nearby galaxy cluster in the early phase of a major
merger. The two-dimensional temperature map using XMM-Newton EPIC-PN
observation confirms the existence of a high temperature region, which we call
the "hot region", in the "bridge region" connecting the two clusters. The ~ 500
kpc wide region between the southeast and northwest boundaries also has higher
pseudo pressure compared to the unshocked regions, suggesting the existence of
two shocks. The southern shock front is clearly visible in the X-ray surface
brightness image and has already been reported by Kato et al. (2015). The
northern one, on the other hand, is newly discovered. To evaluate their Mach
number, we constructed a three-dimensional toy merger model with overlapping
shocked and unshocked components in line of sight. The unshocked and preshock
ICM conditions are estimated based on those outside the interacting bridge
region assuming point symmetry. The hot region spectra are modeled with
two-temperature thermal components, assuming that the shocked condition follows
the Rankin-Hugoniot relation with the preshock condition. As a result, the
shocked region is estimated to have a line-of-sight depth of ~ 1 Mpc with a
Mach number of ~ 1.3 in the southeast shock and ~ 1.7 in the northwest shock.
The age of the shock waves is estimated to be ~ 260 Myr. This three dimensional
merger model is consistent with the Sunyaev-Zeldovich signal obtained using the
Planck observation within the CMB fluctuations. The total flow of the kinetic
energy of the ICM through the southeast shock was estimated to be ~ 2.2 x
$10^{42}$ erg/s. Assuming that 10 % of this energy is converted into ICM
turbulence, the line-of-sight velocity dispersion is calculated to be ~ 200
km/s, which is basically resolvable via coming high spectral resolution
observations.
| arxiv topic:astro-ph.GA astro-ph.HE |
arxiv_dataset-172362210.02245 | Channel Modeling for UAV-to-Ground Communications with Posture Variation
and Fuselage Scattering Effect
eess.SP eess.IV
Unmanned aerial vehicle (UAV)-to-ground (U2G) channel models play a pivotal
role for reliable communications between UAV and ground terminal. This paper
proposes a three-dimensional (3D) non-stationary hybrid model including both
large-scale and small-scale fading for U2G multiple-input-multiple-output
(MIMO) channels. Distinctive channel characteristics under U2G scenarios, i.e.,
3D trajectory and posture of UAV, fuselage scattering effect (FSE), and posture
variation fading (PVF), are incorporated into the proposed model. The channel
parameters, i.e., path loss (PL), shadow fading (SF), path delay, and path
angle, are generated incorporating machine learning (ML) and ray tracing (RT)
techniques to capture the structure-related characteristics. In order to
guarantee the physical continuity of channel parameters such as Doppler phase
and path power, the time evolution methods of inter- and intra- stationary
intervals are proposed. Key statistical properties , i.e., temporal
autocorrection function (ACF), power delay profile (PDP), level crossing rate
(LCR), average fading duration (AFD), and stationary interval (SI) are given,
and the impact of the change of fuselage and posture variation is analyzed. It
is demonstrated that both posture variation and fuselage scattering have
crucial effects on channel characteristics. The validity and practicability of
the proposed model are verified by comparing the simulation results with the
measured ones.
| arxiv topic:eess.SP eess.IV |
arxiv_dataset-172372210.02345 | SCTOMP: Spatially Constrained Time-Optimal Motion Planning
cs.RO math.OC
This paper focuses on spatial time-optimal motion planning, a generalization
of the exact time-optimal path following problem that allows the system to plan
within a predefined space. In contrast to state-of-the-art methods, we drop the
assumption that a collision-free geometric reference is given. Instead, we
present a two-stage motion planning method that solely relies on a goal
location and a geometric representation of the environment to compute a
time-optimal trajectory that is compliant with system dynamics and constraints.
To do so, the proposed scheme first computes an obstacle-free Pythagorean
Hodograph parametric spline, and second solves a spatially reformulated
minimum-time optimization problem. The spline obtained in the first stage is
not a geometric reference, but an extension of the environment representation,
and thus, time-optimality of the solution is guaranteed. The efficacy of the
proposed approach is benchmarked by a known planar example and validated in a
more complex spatial system, illustrating its versatility and applicability.
| arxiv topic:cs.RO math.OC |
arxiv_dataset-172382210.02445 | Localizing Anatomical Landmarks in Ocular Images using Zoom-In Attentive
Networks
eess.IV cs.CV cs.LG
Localizing anatomical landmarks are important tasks in medical image
analysis. However, the landmarks to be localized often lack prominent visual
features. Their locations are elusive and easily confused with the background,
and thus precise localization highly depends on the context formed by their
surrounding areas. In addition, the required precision is usually higher than
segmentation and object detection tasks. Therefore, localization has its unique
challenges different from segmentation or detection. In this paper, we propose
a zoom-in attentive network (ZIAN) for anatomical landmark localization in
ocular images. First, a coarse-to-fine, or "zoom-in" strategy is utilized to
learn the contextualized features in different scales. Then, an attentive
fusion module is adopted to aggregate multi-scale features, which consists of
1) a co-attention network with a multiple regions-of-interest (ROIs) scheme
that learns complementary features from the multiple ROIs, 2) an
attention-based fusion module which integrates the multi-ROIs features and
non-ROI features. We evaluated ZIAN on two open challenge tasks, i.e., the
fovea localization in fundus images and scleral spur localization in AS-OCT
images. Experiments show that ZIAN achieves promising performances and
outperforms state-of-the-art localization methods. The source code and trained
models of ZIAN are available at
https://github.com/leixiaofeng-astar/OMIA9-ZIAN.
| arxiv topic:eess.IV cs.CV cs.LG |
arxiv_dataset-172392210.02545 | JoeyS2T: Minimalistic Speech-to-Text Modeling with JoeyNMT
cs.CL cs.SD eess.AS
JoeyS2T is a JoeyNMT extension for speech-to-text tasks such as automatic
speech recognition and end-to-end speech translation. It inherits the core
philosophy of JoeyNMT, a minimalist NMT toolkit built on PyTorch, seeking
simplicity and accessibility. JoeyS2T's workflow is self-contained, starting
from data pre-processing, over model training and prediction to evaluation, and
is seamlessly integrated into JoeyNMT's compact and simple code base. On top of
JoeyNMT's state-of-the-art Transformer-based encoder-decoder architecture,
JoeyS2T provides speech-oriented components such as convolutional layers,
SpecAugment, CTC-loss, and WER evaluation. Despite its simplicity compared to
prior implementations, JoeyS2T performs competitively on English speech
recognition and English-to-German speech translation benchmarks. The
implementation is accompanied by a walk-through tutorial and available on
https://github.com/may-/joeys2t.
| arxiv topic:cs.CL cs.SD eess.AS |
arxiv_dataset-172402210.02645 | Quantum-improved phase estimation with a displacement-assisted SU(1,1)
interferometer
quant-ph
By performing two local displacement operations (LDOs) inside an SU(1,1)
interferometer, called as the displacement-assisted SU(1,1) [DSU(1,1)], both
the phase sensitivity based on homodyne detection and quantum Fisher
information (QFI) with and without photon losses are investigated in this
paper. In this DSU(1,1) interferometer, we focus our attention on the extent to
which the introduced LDO affects the phase sensitivity and the QFI, even in the
realistic scenario. Our analyses show that the estimation performance of
DSU(1,1) interferometer is always better than that of SU(1,1) interferometer
without the LDO, especially the phase precision of the former in the ideal
scenario gradually approaching to the Heisenberg limit via the increase of the
LDO strength. More significantly, different from the latter, the robustness of
the former can be enhanced markedly by regulating and controlling the LDO. Our
findings would open an useful view for quantum-improved phase estimation of
optical interferometers.
| arxiv topic:quant-ph |
arxiv_dataset-172412210.02745 | MuS2: A Real-World Benchmark for Sentinel-2 Multi-Image Super-Resolution
eess.IV cs.CV
Insufficient image spatial resolution is a serious limitation in many
practical scenarios, especially when acquiring images at a finer scale is
infeasible or brings higher costs. This is inherent to remote sensing,
including Sentinel-2 satellite images that are available free of charge at a
high revisit frequency, but whose spatial resolution is limited to 10 m ground
sampling distance. The resolution can be increased with super-resolution
algorithms, in particular when performed from multiple images captured at
subsequent revisits of a satellite, taking advantage of information fusion that
leads to enhanced reconstruction accuracy. One of the obstacles in multi-image
super-resolution consists in the scarcity of real-world benchmarks - commonly,
simulated data are exploited which do not fully reflect the operating
conditions. In this paper, we introduce a new MuS2 benchmark for
super-resolving multiple Sentinel-2 images, with WorldView-2 imagery used as
the high-resolution reference. Within MuS2, we publish the first end-to-end
evaluation procedure for this problem which we expect to help the researchers
in advancing the state of the art in multi-image super-resolution.
| arxiv topic:eess.IV cs.CV |
arxiv_dataset-172422210.02845 | A unified steady and unsteady formulation for hydrodynamic potential
flow simulations with fully nonlinear free surface boundary conditions
math.NA cs.NA physics.app-ph physics.flu-dyn
This work discusses the correct modeling of the fully nonlinear free surface
boundary conditions to be prescribed in water waves flow simulations based on
potential flow theory. The main goal of such a discussion is that of
identifying a mathematical formulation and a numerical treatment that can be
used both to carry out transient simulations, and to compute steady solutions
-- for any flow admitting them. In the literature on numerical towing tank in
fact, steady and unsteady fully nonlinear potential flow solvers are
characterized by different mathematical formulations. The kinematic and dynamic
fully nonlinear free surface boundary conditions are discussed, and in
particular it is proven that the kinematic free surface boundary condition,
written in semi-Lagrangian form, can be manipulated to derive an alternative
non penetration boundary condition by all means identical to the one used on
the surface of floating bodies or on the basin bottom. The simplified
mathematical problem obtained is discretized over space and time via Boundary
Element Method (BEM) and Implicit Backward Difference Formula (BDF) scheme,
respectively. The results confirm that the solver implemented is able to solve
steady potential flow problems just by eliminating null time derivatives in the
unsteady formulation. Numerical results obtained confirm that the solver
implemented is able to accurately reproduce results of classical steady flow
solvers available in the literature.
| arxiv topic:math.NA cs.NA physics.app-ph physics.flu-dyn |
arxiv_dataset-172432210.02945 | On the optimal pivot path of simplex method for linear programming based
on reinforcement learning
math.OC
Based on the existing pivot rules, the simplex method for linear programming
is not polynomial in the worst case. Therefore the optimal pivot of the simplex
method is crucial. This study proposes the optimal rule to find all shortest
pivot paths of the simplex method for linear programming problems based on
Monte Carlo tree search (MCTS). Specifically, we first propose the
SimplexPseudoTree to transfer the simplex method into tree search mode while
avoiding repeated basis variables. Secondly, we propose four reinforcement
learning (RL) models with two actions and two rewards to make the Monte Carlo
tree search suitable for the simplex method. Thirdly, we set a new action
selection criterion to ameliorate the inaccurate evaluation in the initial
exploration. It is proved that when the number of vertices in the feasible
region is $C_n^m$, our method can generate all the shortest pivot paths, which
is the polynomial of the number of variables. In addition, we experimentally
validate that the proposed schedule can avoid unnecessary search and provide
the optimal pivot path. Furthermore, this method can provide the best pivot
labels for all kinds of supervised learning methods to solve linear programming
problems.
| arxiv topic:math.OC |
arxiv_dataset-172442210.03045 | A neural network approach to high-dimensional optimal switching problems
with jumps in energy markets
math.OC
We develop a backward-in-time machine learning algorithm that uses a sequence
of neural networks to solve optimal switching problems in energy production,
where electricity and fossil fuel prices are subject to stochastic jumps. We
then apply this algorithm to a variety of energy scheduling problems, including
novel high-dimensional energy production problems. Our experimental results
demonstrate that the algorithm performs with accuracy and experiences linear to
sub-linear slowdowns as dimension increases, demonstrating the value of the
algorithm for solving high-dimensional switching problems.
| arxiv topic:math.OC |
arxiv_dataset-172452210.03145 | Snowmass 2021 Topical Report on Synergies in Research at Underground
Facilities
hep-ex
This is a Snowmass 2021 Topical Report for the Underground Facilities and
Infrastructure Frontier on Synergies in Research at Underground Facilities: A
broad range of scientific and engineering research is possible in underground
laboratories, beyond the physics-focused activities described in the other
Underground Facilities and Infrastructure Topical Reports. These areas of
research include nuclear astrophysics, geology, geoengineering, gravitational
wave detection, biology, and perhaps soon quantum information science. This UF
Topical Report will survey those other scientific and engineering research
activities that share interest in research-orientated Underground Facilities
and Infrastructure. In most cases the breadth and depth of research aims is too
large to cover in completeness and references to surveys or key documents for
those fields are provided after introductory summaries. Additional attention is
then given to shared, similar, and unique needs of each research area with
respect to the broader underground research community's Underground Facilities
and Infrastructure needs. Where potential conflicts of usage type, site, or
duration might arise, these are identified.
| arxiv topic:hep-ex |
arxiv_dataset-172462210.03245 | 21st Century Global and Regional Surface Temperature Projections
physics.ao-ph astro-ph.EP physics.soc-ph
Many regions across the globe broke their surface temperature records in
recent years, further sparking concerns about the impending arrival of "tipping
points" later in the 21st century. This study analyzes observed global surface
temperature trends in three target latitudinal regions: the Arctic Circle, the
Tropics, and the Antarctic Circle. We show that global warming is accelerating
unevenly across the planet, with the Arctic warming at approximately three
times the average rate of our world. We further analyzed the reliability of
latitude-dependent surface temperature simulations from a suite of Coupled
Model Intercomparison Project Phase 6 models and their multi-model mean. We
found that GISS-E2-1-G and FGOALS-g3 were the best-performing models based on
their statistical abilities to reproduce observational, latitude-dependent
data. Surface temperatures were projected from ensemble simulations of the
Shared Socioeconomic Pathway 2-4.5 (SSP2-4.5). We estimate when the climate
will warm by 1.5, 2.0, and 2.5 degrees C relative to the preindustrial period,
globally and regionally. GISS-E2-1-G projects that global surface temperature
anomalies would reach 1.5, 2.0, and 2.5 degrees C in 2024 (+/-1.34), 2039
(+/-2.83), and 2057 (+/-5.03) respectively, while FGOALS-g3 predicts these
"tipping points" would arrive in 2024 (+/-2.50), 2054 (+/-7.90), and 2087
(+/-10.55) respectively. Our results reaffirm a dramatic, upward trend in
projected climate warming acceleration, with upward concavity in 21st century
projections of the Arctic, which could lead to catastrophic consequences across
the Earth. Further studies are necessary to determine the most efficient
solutions to reduce global warming acceleration and maintain a low SSP, both
globally and regionally.
| arxiv topic:physics.ao-ph astro-ph.EP physics.soc-ph |
arxiv_dataset-172472210.03345 | Spatial-chirp Codebook-based Hierarchical Beam Training for Extremely
Large-Scale Massive MIMO
cs.IT eess.SP math.IT
Extremely large-scale multiple-input multiple-output (XL-MIMO) promises to
provide ultrahigh data rates in millimeter-wave (mmWave) and Terahertz (THz)
spectrum. However, the spherical-wavefront wireless transmission caused by
large aperture array presents huge challenges for channel state information
(CSI) acquisition and beamforming. Two independent parameters (physical angles
and transmission distance) should be simultaneously considered in XL-MIMO
beamforming, which brings severe overhead consumption and beamforming
degradation. To address this problem, we exploit the near-field channel
characteristic and propose two low-overhead hierarchical beam training schemes
for near-field XL-MIMO system. Firstly, we project near-field channel into
spatial-angular domain and slope-intercept domain to capture detailed
representations. Then we point out three critical criteria for XL-MIMO
hierarchical beam training. Secondly, a novel spatial-chirp beam-aided codebook
and corresponding hierarchical update policy are proposed. Thirdly, given the
imperfect coverage and overlapping of spatial-chirp beams, we further design an
enhanced hierarchical training codebook via manifold optimization and
alternative minimization. Theoretical analyses and numerical simulations are
also displayed to verify the superior performances on beamforming and training
overhead.
| arxiv topic:cs.IT eess.SP math.IT |
arxiv_dataset-172482210.03445 | Experimental and observational tests of antigravity
astro-ph.CO gr-qc hep-ex hep-th
Whereas repulsive gravity was considered as a fringe concept until the
mid-1990's, the growingexperimental evidence since this epoch for repulsive
gravity, in what is now called Dark Energy,for lack of a better understanding
of its nature, has led to a vast literature in order to attemptto characterize
this repulsive component, and notably its equation of state. In the following,
Iwill show that we can use cosmology to test the hypothesis that antimatter is
at the origin ofrepulsive gravity, may play the role of a Dark Energy component
and, more surprisingly, maymimic the presence of Dark Matter, and justify the
MOND phenomenology. More directly,three experiments, AEgIS, ALPHA-g and Gbar,
are attempting to measure the action ofgravitation on cold atoms of
antihydrogen at CERN in a near future. Finally, I note thatCP violation might
be explained by antigravity and I briefly recall the motivations for
thisassertion.
| arxiv topic:astro-ph.CO gr-qc hep-ex hep-th |
arxiv_dataset-172492210.03545 | Hypergraph Ramsey numbers of cliques versus stars
math.CO
Let $K_m^{(3)}$ denote the complete $3$-uniform hypergraph on $m$ vertices
and $S_n^{(3)}$ the $3$-uniform hypergraph on $n+1$ vertices consisting of all
$\binom{n}{2}$ edges incident to a given vertex. Whereas many hypergraph Ramsey
numbers grow either at most polynomially or at least exponentially, we show
that the off-diagonal Ramsey number $r(K_{4}^{(3)},S_n^{(3)})$ exhibits an
unusual intermediate growth rate, namely, \[ 2^{c \log^2 n} \le
r(K_{4}^{(3)},S_n^{(3)}) \le 2^{c' n^{2/3}\log n} \] for some positive
constants $c$ and $c'$. The proof of these bounds brings in a novel Ramsey
problem on grid graphs which may be of independent interest: what is the
minimum $N$ such that any $2$-edge-coloring of the Cartesian product $K_N
\square K_N$ contains either a red rectangle or a blue $K_n$?
| arxiv topic:math.CO |
arxiv_dataset-172502210.03645 | FAST observations of an extremely active episode of FRB 20201124A: II.
Energy Distribution
astro-ph.HE
We report the properties of more than 800 bursts detected from the repeating
fast radio burst (FRB) source FRB 20201124A with the Five-hundred-meter
Aperture Spherical radio Telescope (FAST) during an extremely active episode on
UTC September 25-28, 2021 in a series of four papers. In this second paper of
the series, we mainly focus on the energy distribution of the detected bursts.
The event rate initially increased exponentially but the source activity
stopped within 24 hours after the 4th day. The detection of 542 bursts in one
hour during the fourth day marked the highest event rate detected from one
single FRB source so far. The bursts have complex structures in the
time-frequency space. We find a double-peak distribution of the waiting time,
which can be modeled with two log-normal functions peaking at 51.22 ms and
10.05 s, respectively. Compared with the emission from a previous active
episode of the source detected with FAST, the second distribution peak time is
smaller, suggesting that this peak is defined by the activity level of the
source. We calculate the isotropic energy of the bursts using both a partial
bandwidth and a full bandwidth and find that the energy distribution is not
significantly changed. We find that an exponentially connected broken-power-law
function can fit the cumulative burst energy distribution well, with the lower
and higher-energy indices being $-1.22\pm0.01$ and $-4.27\pm0.23$,
respectively. Assuming a radio radiative efficiency of $\eta_r = 10^{-4}$, the
total isotropic energy of the bursts released during the four days when the
source was active is already $3.9\times10^{46}$ erg, exceeding $\sim 23\%$ of
the available magnetar dipolar magnetic energy. This challenges the magnetar
models invoking an inefficient radio emission (e.g. synchrotron maser models).
| arxiv topic:astro-ph.HE |
arxiv_dataset-172512210.03745 | ProGReST: Prototypical Graph Regression Soft Trees for Molecular
Property Prediction
q-bio.QM cs.AI cs.LG
In this work, we propose the novel Prototypical Graph Regression
Self-explainable Trees (ProGReST) model, which combines prototype learning,
soft decision trees, and Graph Neural Networks. In contrast to other works, our
model can be used to address various challenging tasks, including compound
property prediction. In ProGReST, the rationale is obtained along with
prediction due to the model's built-in interpretability. Additionally, we
introduce a new graph prototype projection to accelerate model training.
Finally, we evaluate PRoGReST on a wide range of chemical datasets for
molecular property prediction and perform in-depth analysis with chemical
experts to evaluate obtained interpretations. Our method achieves competitive
results against state-of-the-art methods.
| arxiv topic:q-bio.QM cs.AI cs.LG |
arxiv_dataset-172522210.03845 | Born this way: thin disc, thick disc, and isotropic spheroid formation
in FIRE-2 Milky-Way-mass galaxy simulations
astro-ph.GA
We investigate the formation of Milky-Way-mass galaxies using FIRE-2 LCDM
cosmological zoom-in simulations by studying the orbital evolution of stars
formed in the main progenitor of the galaxy, from birth to the present day. We
classify in situ stars as isotropic spheroid, thick-disc, and thin-disc
according to their orbital circularities and show that these components are
assembled in a time-ordered sequence from early to late times, respectively.
All simulated galaxies experience an early phase of bursty star formation that
transitions to a late-time steady phase. This transition coincides with the
time that the inner CGM virializes. During the early bursty phase, galaxies
have irregular morphologies and new stars are born on radial orbits; these
stars evolve into an isotropic spheroidal population today. The bulk of
thick-disc stars form at intermediate times, during a clumpy-disc ``spin-up''
phase, slightly later than the peak of spheroid formation. At late times, once
the CGM virializes and star formation ``cools down," stars are born on circular
orbits within a narrow plane. Those stars mostly inhabit thin discs today.
Broadly speaking, stars with disc-like or spheroid-like orbits today were born
that way. Mergers onto discs and secular processes do affect kinematics in our
simulations, but play only secondary roles in populating thick-disc and in situ
spheroid populations at z=0. The age distributions of spheroid, thick disc, and
thin disc populations scale self-similarly with the steady-phase transition
time, which suggests that morphological age dating can be linked to the CGM
virialization time in galaxies.
| arxiv topic:astro-ph.GA |
arxiv_dataset-172532210.03945 | Understanding HTML with Large Language Models
cs.LG cs.AI
Large language models (LLMs) have shown exceptional performance on a variety
of natural language tasks. Yet, their capabilities for HTML understanding --
i.e., parsing the raw HTML of a webpage, with applications to automation of
web-based tasks, crawling, and browser-assisted retrieval -- have not been
fully explored. We contribute HTML understanding models (fine-tuned LLMs) and
an in-depth analysis of their capabilities under three tasks: (i) Semantic
Classification of HTML elements, (ii) Description Generation for HTML inputs,
and (iii) Autonomous Web Navigation of HTML pages. While previous work has
developed dedicated architectures and training procedures for HTML
understanding, we show that LLMs pretrained on standard natural language
corpora transfer remarkably well to HTML understanding tasks. For instance,
fine-tuned LLMs are 12% more accurate at semantic classification compared to
models trained exclusively on the task dataset. Moreover, when fine-tuned on
data from the MiniWoB benchmark, LLMs successfully complete 50% more tasks
using 192x less data compared to the previous best supervised model. Out of the
LLMs we evaluate, we show evidence that T5-based models are ideal due to their
bidirectional encoder-decoder architecture. To promote further research on LLMs
for HTML understanding, we create and open-source a large-scale HTML dataset
distilled and auto-labeled from CommonCrawl.
| arxiv topic:cs.LG cs.AI |
arxiv_dataset-172542210.04045 | The FBHHRBNRSSSHK-Algorithm for Multiplication in
$\mathbb{Z}_2^{5\times5}$ is still not the end of the story
cs.SC cs.CC
In response to a recent Nature article which announced an algorithm for
multiplying $5\times5$-matrices over $\mathbb{Z}_2$ with only 96
multiplications, two fewer than the previous record, we present an algorithm
that does the job with only 95 multiplications.
| arxiv topic:cs.SC cs.CC |
arxiv_dataset-172552210.04145 | Fine-grained Anomaly Detection in Sequential Data via Counterfactual
Explanations
cs.LG
Anomaly detection in sequential data has been studied for a long time because
of its potential in various applications, such as detecting abnormal system
behaviors from log data. Although many approaches can achieve good performance
on anomalous sequence detection, how to identify the anomalous entries in
sequences is still challenging due to a lack of information at the entry-level.
In this work, we propose a novel framework called CFDet for fine-grained
anomalous entry detection. CFDet leverages the idea of interpretable machine
learning. Given a sequence that is detected as anomalous, we can consider
anomalous entry detection as an interpretable machine learning task because
identifying anomalous entries in the sequence is to provide an interpretation
to the detection result. We make use of the deep support vector data
description (Deep SVDD) approach to detect anomalous sequences and propose a
novel counterfactual interpretation-based approach to identify anomalous
entries in the sequences. Experimental results on three datasets show that
CFDet can correctly detect anomalous entries.
| arxiv topic:cs.LG |
arxiv_dataset-172562210.04245 | The distribution of globular clusters in kinematic spaces does not trace
the accretion history of the host galaxy
astro-ph.GA
Reconstructing how all the stellar components of the Galaxy formed and
assembled over time, by studying the properties of the stars which make it, is
the aim of Galactic archeology. In these last years, thanks to the launch of
the ESA Gaia astrometric mission, and the development of many spectroscopic
surveys, we are for the first time in the position to delve into the layers of
the past of our galaxy. Globular clusters (GCs) play a fundamental role in this
research field since they are among the oldest stellar systems in the Milky Way
(MW) and so bear witness of its entire past. In the recent years, there have
been several attempts to constrain the nature of clusters (accreted or formed
in the MW itself) through the analysis of kinematic spaces and to reconstruct
from this the properties of the accretions events experienced by the MW through
time. This work aims to test a widely-used assumption about the clustering of
the accreted populations of GCs in the integrals of motions space. We analyze a
set of dissipation-less N-body simulations that reproduce the accretion of one
or two satellites with their GC population on a MW-type galaxy. Our results
demonstrate that a significant overlap between accreted and
"kinematically-heated" in-situ GCs is expected in kinematic spaces, for mergers
with mass ratios of 1:10. In contrast with standard assumptions made in the
literature so far, we find that accreted GCs do not show dynamical coherence,
that is they do not cluster in kinematic spaces. In addition, GCs can also be
found in regions dominated by stars which have a different origin (i.e.
different progenitor). This casts doubt on the association between GCs and
field stars that is generally made in the literature to assign them to a common
origin. Our findings severely question the recovered accretion history of the
MW based on the phase-space clustering of the GC population.
| arxiv topic:astro-ph.GA |
arxiv_dataset-172572210.04345 | LieGG: Studying Learned Lie Group Generators
cs.LG
Symmetries built into a neural network have appeared to be very beneficial
for a wide range of tasks as it saves the data to learn them. We depart from
the position that when symmetries are not built into a model a priori, it is
advantageous for robust networks to learn symmetries directly from the data to
fit a task function. In this paper, we present a method to extract symmetries
learned by a neural network and to evaluate the degree to which a network is
invariant to them. With our method, we are able to explicitly retrieve learned
invariances in a form of the generators of corresponding Lie-groups without
prior knowledge of symmetries in the data. We use the proposed method to study
how symmetrical properties depend on a neural network's parameterization and
configuration. We found that the ability of a network to learn symmetries
generalizes over a range of architectures. However, the quality of learned
symmetries depends on the depth and the number of parameters.
| arxiv topic:cs.LG |
arxiv_dataset-172582210.04445 | Enumerating Stable Nanopores in Graphene and their Geometrical
Properties Using the Combinatorics of Hexagonal Lattices
cond-mat.mtrl-sci physics.data-an
Nanopores in two-dimensional (2D) materials, including graphene, can be used
for a variety of applications, such as gas separations, water desalination, and
DNA sequencing. So far, however, all plausible isomeric shapes of graphene
nanopores have not been enumerated. Instead, a probabilistic approach has been
followed to predict nanopore shapes in 2D materials, due to the exponential
increase in the number of nanopores as the size of the vacancy increases. For
example, there are 12 possible isomers when N=6 atoms are removed, a number
that theoretically increases to 11.7 million when N=20 atoms are removed from
the graphene lattice. The development of a smaller, exhaustive dataset of
stable nanopore shapes can help future experimental and theoretical studies
focused on using nanoporous 2D materials in various applications. In this work,
we use the theory of 2D triangular "lattice animals" to create a library of all
stable graphene nanopore shapes based on a modification of a well-known
algorithm in the mathematical combinatorics of polyforms known as Redelmeier's
algorithm. We show that there exists a correspondence between graphene
nanopores and triangular polyforms (called polyiamonds) as well as hexagonal
polyforms (called polyhexes). We develop the concept of a polyiamond ID to
identify unique nanopore isomers. We also use concepts from polyiamond and
polyhex geometry to eliminate unstable nanopores containing dangling atoms,
bonds, and moieties. The exclusion of such unstable nanopores leads to a
remarkable reduction in the possible nanopores from 11.7 million for N=20 to
only 0.184 million nanopores, thereby indicating that the number of stable
nanopores is almost two orders of magnitude lower and is much more tractable.
Not only that, by extracting the polyhex outline, our algorithm allows
searching for nanopores with dimensions and shape factors in a specified range.
| arxiv topic:cond-mat.mtrl-sci physics.data-an |
arxiv_dataset-172592210.04545 | Automatic Evaluation and Analysis of Idioms in Neural Machine
Translation
cs.CL
A major open problem in neural machine translation (NMT) is the translation
of idiomatic expressions, such as "under the weather". The meaning of these
expressions is not composed by the meaning of their constituent words, and NMT
models tend to translate them literally (i.e., word-by-word), which leads to
confusing and nonsensical translations. Research on idioms in NMT is limited
and obstructed by the absence of automatic methods for quantifying these
errors. In this work, first, we propose a novel metric for automatically
measuring the frequency of literal translation errors without human
involvement. Equipped with this metric, we present controlled translation
experiments with models trained in different conditions (with/without the
test-set idioms) and across a wide range of (global and targeted) metrics and
test sets. We explore the role of monolingual pretraining and find that it
yields substantial targeted improvements, even without observing any
translation examples of the test-set idioms. In our analysis, we probe the role
of idiom context. We find that the randomly initialized models are more local
or "myopic" as they are relatively unaffected by variations of the idiom
context, unlike the pretrained ones.
| arxiv topic:cs.CL |
arxiv_dataset-172602210.04645 | Optimal Stopping with Trees: The Details
math.OC
The purpose of this paper is two-fold, first, to review a recent method
introduced by S. Becker, P. Cheridito, and P. Jentzen, for solving
high-dimensional optimal stopping problems using deep Neural Networks, second,
to propose an alternative algorithm replacing Neural Networks by CART-trees
which allow for more interpretation of the estimated stopping rules. We in
particular compare the performance of the two algorithms with respect to the
Bermudan max-call benchmark example concluding that the Bermudan max-call may
not be suitable to serve as a benchmark example for high-dimensional optimal
stopping problems. We also show how our algorithm can be used to plot stopping
boundaries.
| arxiv topic:math.OC |
arxiv_dataset-172612210.04745 | Low-power multi-mode fiber projector overcomes shallow neural networks
classifiers
physics.optics physics.app-ph physics.data-an stat.ML
In the domain of disordered photonics, the characterization of optically
opaque materials for light manipulation and imaging is a primary aim. Among
various complex devices, multi-mode optical fibers stand out as cost-effective
and easy-to-handle tools, making them attractive for several tasks. In this
context, we cast these fibers into random hardware projectors, transforming an
input dataset into a higher dimensional speckled image set. The goal of our
study is to demonstrate that using such randomized data for classification by
training a single logistic regression layer improves accuracy compared to
training on direct raw images. Interestingly, we found that the classification
accuracy achieved is higher than that obtained with the standard transmission
matrix model, a widely accepted tool for describing light transmission through
disordered devices. We conjecture that the reason for such improved performance
could be due to the fact that the hardware classifier operates in a flatter
region of the loss landscape when trained on fiber data, which aligns with the
current theory of deep neural networks. These findings suggest that the class
of random projections operated by multi-mode fibers generalize better to
previously unseen data, positioning them as promising tools for
optically-assisted neural networks. With this study, in fact, we want to
contribute to advancing the knowledge and practical utilization of these
versatile instruments, which may play a significant role in shaping the future
of neuromorphic machine learning.
| arxiv topic:physics.optics physics.app-ph physics.data-an stat.ML |
arxiv_dataset-172622210.04845 | FS-DETR: Few-Shot DEtection TRansformer with prompting and without
re-training
cs.CV cs.AI
This paper is on Few-Shot Object Detection (FSOD), where given a few
templates (examples) depicting a novel class (not seen during training), the
goal is to detect all of its occurrences within a set of images. From a
practical perspective, an FSOD system must fulfil the following desiderata: (a)
it must be used as is, without requiring any fine-tuning at test time, (b) it
must be able to process an arbitrary number of novel objects concurrently while
supporting an arbitrary number of examples from each class and (c) it must
achieve accuracy comparable to a closed system. Towards satisfying (a)-(c), in
this work, we make the following contributions: We introduce, for the first
time, a simple, yet powerful, few-shot detection transformer (FS-DETR) based on
visual prompting that can address both desiderata (a) and (b). Our system
builds upon the DETR framework, extending it based on two key ideas: (1) feed
the provided visual templates of the novel classes as visual prompts during
test time, and (2) ``stamp'' these prompts with pseudo-class embeddings (akin
to soft prompting), which are then predicted at the output of the decoder.
Importantly, we show that our system is not only more flexible than existing
methods, but also, it makes a step towards satisfying desideratum (c).
Specifically, it is significantly more accurate than all methods that do not
require fine-tuning and even matches and outperforms the current
state-of-the-art fine-tuning based methods on the most well-established
benchmarks (PASCAL VOC & MSCOCO).
| arxiv topic:cs.CV cs.AI |
arxiv_dataset-172632210.04945 | Spectroscopic Observation and Modeling of Photonic Modes in CeO2
Nanocubes
physics.optics cond-mat.mtrl-sci
Photonic modes in dielectric nanostructures, e.g., wide gap semiconductor
like CeO2 (ceria), has potential for various applications such as light
harvesting and information transmission. To fully understand the properties of
such phenomenon in nanoscale, we applied electron energy-loss spectroscopy
(EELS) in scanning transmission electron microscope (STEM) to detect such modes
in a well-defined ceria nanocube. Through spectra and mapping, we demonstrated
a geometrical difference of mode excitation. By comparing various spectra taken
at different location relative to the cube, we also showed the transmission
properties of the mode. To confirm our observation, we performed EELS
simulation with finite-element dielectric calculations in COMSOL Multiphysics.
We also revealed the origin of the modes through the calculation. We purposed a
simple analytical model to estimate the energy of photonic modes as well. In
all, this work gave a fine description of the photonic modes' properties in
nanostructures, while demonstrating the advantage of EELS in characterizing
optical phenomena in nanoscale.
| arxiv topic:physics.optics cond-mat.mtrl-sci |
arxiv_dataset-172642210.05045 | Formation of amorphous carbon multi-walled nanotubes from random initial
configurations
cond-mat.mtrl-sci cond-mat.dis-nn
Amorphous carbon nanotubes (a-CNT) with up to four walls and sizes ranging
from 200 to 3200 atoms have been simulated, starting from initial random
configurations and using the Gaussian Approximation Potential [Phys. Rev. B 95,
094203 (2017)]. The important variables (like density, height, and diameter)
required to successfully simulate a-CNTs, were predicted with a machine
learning random forest technique. The models were validated $\textit{ex post
facto}$ using density functional codes. The a-CNT models ranged from 0.55 nm -
2 nm wide with an average inter-wall spacing of 0.31 nm. The topological
defects in a-CNTs were discussed and new defect configurations were observed.
The electronic density of states and localization in these phases were
discussed and delocalized electrons in the $\pi$ subspace were identified as an
important factor for inter-layer cohesion. Spatial projection of the electronic
conductivity favors axial transport along connecting hexagons, while
non-hexagonal parts of the network either hinder or bifurcate the electronic
transport. A vibrational density of states was calculated and is potentially an
experimentally testable fingerprint of the material and the appearance of a
low-frequency radial breathing mode was discussed. The thermal conductivity at
300 K was calculated using the Green-Kubo formula.
| arxiv topic:cond-mat.mtrl-sci cond-mat.dis-nn |
arxiv_dataset-172652210.05145 | Retrieval Augmentation for T5 Re-ranker using External Sources
cs.IR cs.CL
Retrieval augmentation has shown promising improvements in different tasks.
However, whether such augmentation can assist a large language model based
re-ranker remains unclear. We investigate how to augment T5-based re-rankers
using high-quality information retrieved from two external corpora -- a
commercial web search engine and Wikipedia. We empirically demonstrate how
retrieval augmentation can substantially improve the effectiveness of T5-based
re-rankers for both in-domain and zero-shot out-of-domain re-ranking tasks.
| arxiv topic:cs.IR cs.CL |
arxiv_dataset-172662210.05245 | PatternRank: Leveraging Pretrained Language Models and Part of Speech
for Unsupervised Keyphrase Extraction
cs.CL cs.AI
Keyphrase extraction is the process of automatically selecting a small set of
most relevant phrases from a given text. Supervised keyphrase extraction
approaches need large amounts of labeled training data and perform poorly
outside the domain of the training data. In this paper, we present PatternRank,
which leverages pretrained language models and part-of-speech for unsupervised
keyphrase extraction from single documents. Our experiments show PatternRank
achieves higher precision, recall and F1-scores than previous state-of-the-art
approaches. In addition, we present the KeyphraseVectorizers package, which
allows easy modification of part-of-speech patterns for candidate keyphrase
selection, and hence adaptation of our approach to any domain.
| arxiv topic:cs.CL cs.AI |
arxiv_dataset-172672210.05345 | Sympletic tracking methods for insertion devices: a Robinson wiggler
example
physics.acc-ph
Modern synchrotron light sources are often characterized with high-brightness
synchrotron radiation from insertion devices. Inevitably, insertion devices
introduce nonlinear distortion to the beam motion. Symplectic tracking is
crucial to study the impact, especially for the low- and medium-energy storage
rings. This paper uses a Robinson wiggler as an example to illustrate an
universally applicable analytical representation of the magnetic field and to
summarizes four different symplectic tracking methods.
| arxiv topic:physics.acc-ph |
arxiv_dataset-172682210.05445 | Borel $(\alpha,\beta)$-multitransforms and Quantum Leray-Hirsch:
integral representations of solutions of quantum differential equations for
$\mathbb P^1$-bundles
math.AG math-ph math.DG math.MP
In this paper, we address the integration problem of the isomonodromic system
of quantum differential equations ($qDE$s) associated with the quantum
cohomology of $\mathbb P^1$-bundles on Fano varieties. It is shown that bases
of solutions of the $qDE$ of the total space of the $\mathbb P^1$-bundle can be
reconstructed from the datum of bases of solutions of the corresponding $qDE$
associated with the base space. This represents a quantum analog of the
classical Leray-Hirsch theorem in the context of the isomonodromic approach to
quantum cohomology. The reconstruction procedure of the solutions can be
performed in terms of some integral transforms, introduced in arXiv:2005.08262,
called $Borel$ $(\alpha,\beta)$-$multitransf\!orms$. We emphasize the
emergence, in the explicit integral formulas, of an interesting sequence of
special functions (closely related to iterated partial derivatives of the
B\"ohmer-Tricomi incomplete Gamma function) as integral kernels. Remarkably,
these integral kernels have a universal feature, being independent of the
specifically chosen $\mathbb P^1$-bundle. When applied to projective bundles on
products of projective spaces, our results give Mellin-Barnes integral
representations of solutions of $qDE$s. As an example, we show how to integrate
the $qDE$ of blow-up of $\mathbb P^2$ at one point via Borel multitransforms of
solutions of the $qDE$ of $\mathbb P^1$.
| arxiv topic:math.AG math-ph math.DG math.MP |
arxiv_dataset-172692210.05545 | Reliability and Reproducibility of the Cryogenic Sapphire Oscillator
Technology
physics.ins-det
The cryogenic sapphire oscillator (CSO) is a highly specialized machine,
which delivers a reference signal exhibiting the lowest frequency fluctuations.
For the best units, the Allan deviation (ADEV) is <1e-15 for integration time
between 1 and 10,000 s, with a drift <1e-14 in one day.The oscillator is based
on a sapphire monocrystal resonating at 10 GHz in a whispering-gallery mode,
cooled at 6 K for highest Q-factor and zero thermal coefficient. We report on
the progress accomplished implementing eleven CSOs in about 10 years since the
first sample delivered to the ESA station in Argentina.
Short-term stability is improved by a factor of 3-10, depending on the
integration time, and the refrigerator's electric power is reduced to 3 kW.
Frequency stability and overall performances are reproducible, with unattended
operation between scheduled maintenance every two years. The CSO is suitable to
scientific applications requiring extreme frequency stability with reliable
long-term operation. For example, the flywheel for primary frequency standards,
the ground segment of GNSS, astrometry, VLBI, and radio astronomy stations.
| arxiv topic:physics.ins-det |
arxiv_dataset-172702210.05645 | Zero Energy Self-Similar Solutions Describing Singularity Formation In
The Nonlinear Schrodinger Equation In Dimension N=3
math.AP math.CA
In dimension N=3 the cubic nonlinear Schrodinger equation has solutions which
become singular, i.e. at a spatial point they blow up to infinity in finite
time. In 1972 Zakharov famously investigated finite time singularity formation
in the cubic nonlinear Schrodinger equation as a model for spatial collapse of
Langmuir waves in plasma, the most abundant form of observed matter in the
universe. Zakharov assumed that (NLS) blow up of solutions is self-similar and
radially symmetric, and that singularity formation can be modeled by a solution
of an associated self-similar, complex ordinary differential equation~(ODE). A
parameter a>0 appears in the ODE, and the dependent variable, Q, satisfies
(Q(0),Q'(0))=(Q_{0},0), where Q(0)>0. A fundamentally important step towards
putting the Zakharov model on a firm mathematical footing is to prove, when
N=3, whether values a>0 and Q_{0}>0 exist such that Q also satisfies the
physically important `zero-energy' integral constraint. Since 1972 this has
remained an open problem. Here, we resolve this issue by proving that for every
a>0 and Q(0)>0, Q satisfies the the `zero-energy' integral constraint.
| arxiv topic:math.AP math.CA |
arxiv_dataset-172712210.05745 | The science of ultra-high energy cosmic rays after more than 15 years of
operation of the Pierre Auger Observatory
astro-ph.HE
The Pierre Auger Observatory has been detecting ultra-high energy cosmic rays
(UHECRs) for more than fifteen years. An essential feature of the Observatory
is its hybrid design: cosmic rays above $100~$PeV are detected through the
observation of the associated air showers with different and complementary
techniques, from surface detector arrays and fluorescence telescopes to radio
antennas. The analyses of the multi-detector data have enabled high-statistics
and high-precision studies of the energy spectrum, mass composition and
distribution of arrival directions of UHECRs. The resulting picture is
summarized in this contribution. While no discrete source of UHECRs has been
identified so far, the extragalactic origin of the particles has been recently
determined from the arrival directions above 8~EeV, and the ring is closing
around nearby astrophysical sites at higher energies. Also, the established
upper limits on fluxes of UHE neutrinos and photons have implications on dark
matter and cosmological aspects that are also presented in this contribution.
| arxiv topic:astro-ph.HE |
arxiv_dataset-172722210.05845 | Contrastive Retrospection: honing in on critical steps for rapid
learning and generalization in RL
cs.LG cs.AI
In real life, success is often contingent upon multiple critical steps that
are distant in time from each other and from the final reward. These critical
steps are challenging to identify with traditional reinforcement learning (RL)
methods that rely on the Bellman equation for credit assignment. Here, we
present a new RL algorithm that uses offline contrastive learning to hone in on
these critical steps. This algorithm, which we call Contrastive Retrospection
(ConSpec), can be added to any existing RL algorithm. ConSpec learns a set of
prototypes for the critical steps in a task by a novel contrastive loss and
delivers an intrinsic reward when the current state matches one of the
prototypes. The prototypes in ConSpec provide two key benefits for credit
assignment: (i) They enable rapid identification of all the critical steps.
(ii) They do so in a readily interpretable manner, enabling out-of-distribution
generalization when sensory features are altered. Distinct from other
contemporary RL approaches to credit assignment, ConSpec takes advantage of the
fact that it is easier to retrospectively identify the small set of steps that
success is contingent upon (and ignoring other states) than it is to
prospectively predict reward at every taken step. ConSpec greatly improves
learning in a diverse set of RL tasks. The code is available at the link:
https://github.com/sunchipsster1/ConSpec
| arxiv topic:cs.LG cs.AI |
arxiv_dataset-172732210.05945 | On polynomial invariant rings in modular invariant theory
math.AC
Let $\Bbbk$ be a field of characteristic $p>0$, $V$ a finite-dimensional
$\Bbbk$-vector-space, and $G$ a finite $p$-group acting $\Bbbk$-linearly on
$V$. Let $S = \Sym V^*$. We show that $S^G$ is a polynomial ring if and only if
the dimension of its singular locus is less than $\rank_\Bbbk V^G$. Confirming
a conjecture of Shank-Wehlau-Broer, we show that if $S^G$ is a direct summand
of $S$, then $S^G$ is a polynomial ring, in the following cases:
\begin{enumerate}
\item $\Bbbk = \bbF_p$ and $\rank_\Bbbk V^G = 4$; or
\item $|G| = p^3$. \end{enumerate} In order to prove the above result, we
also show that if $\rank_\Bbbk V^G \geq \rank_\Bbbk V - 2$, then the Hilbert
ideal $\hilbertIdeal_{G,S}$ is a complete intersection.
| arxiv topic:math.AC |
arxiv_dataset-172742210.06045 | Multiple measurements on an uncollapsed entangled two-photon state
quant-ph
The relativity of simultaneity together with definition of a quantum state's
collapse result into experimental situations, where a multiple measurements can
be taken on an uncollapsed quantum state. A quantum state's collapse space-time
distribution is shown in a reference frame of a detector measuring the quantum
system and in a reference frame moving relative to the detector. From their
inspection follows, that under certain conditions, multiple measurements are
allowed on the same uncollapsed quantum state. An application of the developed
approach is shown on measurement of photon-pair state entangled in
polarizations and energy. I derive conditions, under which two measurements can
be taken on the uncollapsed photon-pair state. From allowance of multiple
measurements on the same uncollapsed state follow serious consequences. For
example, the measurements taken by both detectors in this situation are
uncorrelated. Moreover, all the conservation laws could be violated in
individual measurements, but not in mean value. This statement is proved on the
two-photon state entangled in energy. This is in contradiction with
experimental results observed by the detectors in rest relative to each other.
If experimental results of the proposed experiment disagree with the
predictions -- results measured on the quantum state are correlated, new
space-time distribution of the quantum state's collapse, different from the
Copenhagen interpretation, has to be proposed for proper solution of this
situation.
| arxiv topic:quant-ph |
arxiv_dataset-172752210.06145 | Can Artificial Intelligence Reconstruct Ancient Mosaics?
cs.CV cs.AI
A large number of ancient mosaics have not reached us because they have been
destroyed by erosion, earthquakes, looting or even used as materials in newer
construction. To make things worse, among the small fraction of mosaics that we
have been able to recover, many are damaged or incomplete. Therefore,
restoration and reconstruction of mosaics play a fundamental role to preserve
cultural heritage and to understand the role of mosaics in ancient cultures.
This reconstruction has traditionally been done manually and more recently
using computer graphics programs but always by humans. In the last years,
Artificial Intelligence (AI) has made impressive progress in the generation of
images from text descriptions and reference images. State of the art AI tools
such as DALL-E2 can generate high quality images from text prompts and can take
a reference image to guide the process. In august 2022, DALL-E2 launched a new
feature called outpainting that takes as input an incomplete image and a text
prompt and then generates a complete image filling the missing parts. In this
paper, we explore whether this innovative technology can be used to reconstruct
mosaics with missing parts. Hence a set of ancient mosaics have been used and
reconstructed using DALL-E2; results are promising showing that AI is able to
interpret the key features of the mosaics and is able to produce
reconstructions that capture the essence of the scene. However, in some cases
AI fails to reproduce some details, geometric forms or introduces elements that
are not consistent with the rest of the mosaic. This suggests that as AI image
generation technology matures in the next few years, it could be a valuable
tool for mosaic reconstruction going forward.
| arxiv topic:cs.CV cs.AI |
arxiv_dataset-172762210.06245 | Back to the Future: On Potential Histories in NLP
cs.CL
Machine learning and NLP require the construction of datasets to train and
fine-tune models. In this context, previous work has demonstrated the
sensitivity of these data sets. For instance, potential societal biases in this
data are likely to be encoded and to be amplified in the models we deploy. In
this work, we draw from developments in the field of history and take a novel
perspective on these problems: considering datasets and models through the lens
of historical fiction surfaces their political nature, and affords
re-configuring how we view the past, such that marginalized discourses are
surfaced. Building on such insights, we argue that contemporary methods for
machine learning are prejudiced towards dominant and hegemonic histories.
Employing the example of neopronouns, we show that by surfacing marginalized
histories within contemporary conditions, we can create models that better
represent the lived realities of traditionally marginalized and excluded
communities.
| arxiv topic:cs.CL |
arxiv_dataset-172772210.06345 | Variational Open-Domain Question Answering
cs.CL cs.IR cs.LG
Retrieval-augmented models have proven to be effective in natural language
processing tasks, yet there remains a lack of research on their optimization
using variational inference. We introduce the Variational Open-Domain (VOD)
framework for end-to-end training and evaluation of retrieval-augmented models,
focusing on open-domain question answering and language modelling. The VOD
objective, a self-normalized estimate of the R\'enyi variational bound,
approximates the task marginal likelihood and is evaluated under samples drawn
from an auxiliary sampling distribution (cached retriever and/or approximate
posterior). It remains tractable, even for retriever distributions defined on
large corpora. We demonstrate VOD's versatility by training reader-retriever
BERT-sized models on multiple-choice medical exam questions. On the MedMCQA
dataset, we outperform the domain-tuned Med-PaLM by +5.3% despite using
2.500$\times$ fewer parameters. Our retrieval-augmented BioLinkBERT model
scored 62.9% on the MedMCQA and 55.0% on the MedQA-USMLE. Last, we show the
effectiveness of our learned retriever component in the context of medical
semantic search.
| arxiv topic:cs.CL cs.IR cs.LG |
arxiv_dataset-172782210.06445 | Hydrodynamic Entropy and Emergence of Order in Two-dimensional Euler
Turbulence
cond-mat.stat-mech nlin.CD physics.flu-dyn physics.plasm-ph
Using numerical simulations, we show that the asymptotic states of
two-dimensional (2D) Euler turbulence exhibit large-scale flow structures due
to nonzero energy transfers among small wavenumber modes. These asymptotic
states, which depend on the initial conditions, are out of equilibrium, and
they are different from the predictions of Onsager and Kraichnan. We propose
``hydrodynamic entropy'' to quantify order in 2D Euler turbulence; we show that
this entropy decreases with time, even though the system is isolated with no
dissipation and no contact with a heat bath.
| arxiv topic:cond-mat.stat-mech nlin.CD physics.flu-dyn physics.plasm-ph |
arxiv_dataset-172792210.06545 | GULP: a prediction-based metric between representations
cs.LG
Comparing the representations learned by different neural networks has
recently emerged as a key tool to understand various architectures and
ultimately optimize them. In this work, we introduce GULP, a family of distance
measures between representations that is explicitly motivated by downstream
predictive tasks. By construction, GULP provides uniform control over the
difference in prediction performance between two representations, with respect
to regularized linear prediction tasks. Moreover, it satisfies several
desirable structural properties, such as the triangle inequality and invariance
under orthogonal transformations, and thus lends itself to data embedding and
visualization. We extensively evaluate GULP relative to other methods, and
demonstrate that it correctly differentiates between architecture families,
converges over the course of training, and captures generalization performance
on downstream linear tasks.
| arxiv topic:cs.LG |
arxiv_dataset-172802210.06645 | Serre curves relative to obstructions modulo 2
math.NT
We consider elliptic curves $E / \mathbb{Q}$ for which the image of the
adelic Galois representation $\rho_E$ is as large as possible given a
constraint on the image modulo 2. For such curves, we give a characterization
in terms of their $\ell$-adic images, compute all examples of conductor at most
500,000, precisely describe the image of $\rho_E$, and offer an application to
the cyclicity problem. In this way, we generalize some foundational results on
Serre curves.
| arxiv topic:math.NT |
arxiv_dataset-172812210.06745 | Insertion of vertex operators using BV formalism
hep-th
We develop a general framework for the insertion of vertex operator on the
string worldsheet, in BV formalism. Such insertions correspond to deformations
of the Master Action which breaks the gauge symmetry to a subgroup, and then
restoring the full gauge symmetry by integrating over a cycle in the space of
Lagrangian submanifolds. We provide the general construction, global on the
moduli space, which was previously conjectured in a form local on the
worldsheet. We explain how the enhancement of the gauge symmetry in equivariant
BV formalism can be seen as an application of the general idea of BV effective
action. We derive an integral formula for the deformation of the contraction
operator due to the vertex insertion.
| arxiv topic:hep-th |
arxiv_dataset-172822210.06845 | The Fine-Grained Complexity of Graph Homomorphism Parameterized by
Clique-Width
cs.CC
The generic homomorphism problem, which asks whether an input graph $G$
admits a homomorphism into a fixed target graph $H$, has been widely studied in
the literature. In this article, we provide a fine-grained complexity
classification of the running time of the homomorphism problem with respect to
the clique-width of $G$ (denoted $\operatorname{cw}$) for virtually all choices
of $H$ under the Strong Exponential Time Hypothesis. In particular, we identify
a property of $H$ called the signature number $s(H)$ and show that for each
$H$, the homomorphism problem can be solved in time
$\mathcal{O}^*(s(H)^{\operatorname{cw}})$. Crucially, we then show that this
algorithm can be used to obtain essentially tight upper bounds. Specifically,
we provide a reduction that yields matching lower bounds for each $H$ that is
either a projective core or a graph admitting a factorization with additional
properties -- allowing us to cover all possible target graphs under
long-standing conjectures.
| arxiv topic:cs.CC |
arxiv_dataset-172832210.06945 | Design and characterization of cochlea-inspired tonotopic resonators
physics.app-ph
The cochlea has long been the subject of investigation in various research
fields due to its intriguing spiral architecture and unique sensing
characteristics. One of its most interesting features is the ability to sense
acoustic waves at different spatial locations, based on their frequency
content. In this work, we propose a novel design for a tonotopic resonator,
based on a cochlea-inspired spiral. The resulting structure was subjected to an
optimization process to exhibit out-of-plane vibration modes with mean
out-of-plane displacement maxima distributed along its centerline spanning
nearly a two-decade frequency range. Numerical simulations are performed to
demonstrate the concept, which is also confirmed experimentally on a 3D printed
structure. The obtained frequency-dependent distribution is shown to be a
viable source of information for the discrimination of signals with various
frequency components. The harnessed tonotopic features can be used as a
fundamental principle to design structures with applications in areas such as
non-destructive testing and vibration attenuation.
| arxiv topic:physics.app-ph |
arxiv_dataset-172842210.07045 | Enlargement of Filtrations -- A Primer
math.PR
In stochastic analysis, the flow of information through time is typically
modelled using a filtration. We introduce some of the basic ideas involving
enlargements of filtration. Here, we focus mainly on initial enlargements,
where a given filtration is enlarged with knowledge of an additional random
variable. This has applications to the modelling of insider trading in
mathematical finance.
| arxiv topic:math.PR |
arxiv_dataset-172852210.07145 | Accurate, reliable and interpretable solubility prediction of druglike
molecules with attention pooling and Bayesian learning
q-bio.BM cs.AI cs.LG
In drug discovery, aqueous solubility is an important pharmacokinetic
property which affects absorption and assay availability of drug. Thus, in
silico prediction of solubility has been studied for its utility in virtual
screening and lead optimization. Recently, machine learning (ML) methods using
experimental data has been popular because physics-based methods like quantum
mechanics and molecular dynamics are not suitable for high-throughput tasks due
to its computational costs. However, ML method can exhibit over-fitting problem
in a data-deficient condition, and this is the case for most chemical property
datasets. In addition, ML methods are regarded as a black box function in that
it is difficult to interpret contribution of hidden features to outputs,
hindering analysis and modification of structure-activity relationship. To deal
with mentioned issues, we developed Bayesian graph neural networks (GNNs) with
the self-attention readout layer. Unlike most GNNs using self-attention in node
updates, self-attention applied at readout layer enabled a model to improve
prediction performance as well as to identify atom-wise importance, which can
help lead optimization as exemplified for three FDA-approved drugs. Also,
Bayesian inference enables us to separate more or less accurate results
according to uncertainty in solubility prediction task We expect that our
accurate, reliable and interpretable model can be used for more careful
decision-making and various applications in the development of drugs.
| arxiv topic:q-bio.BM cs.AI cs.LG |
arxiv_dataset-172862210.07245 | Autoencoder-Aided Visualization of Collections of Morse Complexes
cs.HC cs.GR
Though analyzing a single scalar field using Morse complexes is well studied,
there are few techniques for visualizing a collection of Morse complexes. We
focus on analyses that are enabled by looking at a Morse complex as an embedded
domain decomposition. Specifically, we target 2D scalar fields, and we encode
the Morse complex through binary images of the boundaries of decomposition.
Then we use image-based autoencoders to create a feature space for the Morse
complexes. We apply additional dimensionality reduction methods to construct a
scatterplot as a visual interface of the feature space. This allows us to
investigate individual Morse complexes, as they relate to the collection,
through interaction with the scatterplot. We demonstrate our approach using a
synthetic data set, microscopy images, and time-varying vorticity magnitude
fields of flow. Through these, we show that our method can produce insights
about structures within the collection of Morse complexes.
| arxiv topic:cs.HC cs.GR |
arxiv_dataset-172872210.07345 | A Topological Data Analysis Study on Murine Pulmonary Arterial Trees
with Pulmonary Hypertension
q-bio.QM
Pulmonary hypertension (PH), defined by a mean pulmonary arterial blood
pressure above 20 mmHg, is a cardiovascular disease impacting the pulmonary
vasculature. PH is accompanied by vascular remodeling, wherein vessels become
stiffer, large vessels dilate, and smaller vessels constrict. Some types of PH,
including hypoxia-induced PH (HPH), lead to microvascular rarefaction. The goal
of this study is to analyze the change in pulmonary arterial network
morphometry in the presence of HPH. To do so, we use novel methods from
topological data analysis (TDA), employing persistent homology to quantify
arterial network morphometry for control and hypertensive mice. These methods
are used to characterize arterial trees extracted from micro-computed
tomography (micro-CT) images. To compare results between control and
hypertensive animals, we normalize generated networks using three pruning
algorithms. This proof-of-concept study shows that the pruning methods effects
the spatial tree statistics and complexities of the trees. Results show that
HPH trees have higher depth and that the directional complexities correlate
with branch number, except for trees pruned by vessel radius, where the left
and anterior complexity are lower compared to control trees. While more data is
required to make a conclusion about the overall effect of HPH on network
topology, this study provides a framework for analyzing the topology of
biological networks and is a step towards the extraction of relevant
information for diagnosing and detecting HPH.
| arxiv topic:q-bio.QM |
arxiv_dataset-172882210.07445 | An efficient low-density grating setup for monochromatization of XUV
ultrafast light sources
physics.optics
Ultrafast light sources have become an indispensable tool to access and
understand transient phenomenon in material science. However, a simple and
easy-to-implement method for harmonic selection, with high transmission
efficiency and pulse duration conservation, is still a challenge. Here we
showcase and compare two approaches for selecting the desired harmonic from a
high harmonic generation source while achieving the above goals. The first
approach is the combination of extreme ultraviolet spherical mirrors with
transmission filters and the second approach uses a normal-incidence spherical
grating. Both solutions target time- and angle-resolved photoemission
spectroscopy with photon energies in the 10-20 eV range but are relevant for
other experimental techniques as well. The two approaches for harmonic
selection are characterized in terms of focusing quality, efficiency, and
temporal broadening. It is demonstrated that a focusing grating is able to
provide much higher transmission as compared to the mirror+filter approach (3.3
times higher for 10.8 eV and 12.9 times higher for 18.1 eV), with only a slight
temporal broadening (6.8% increase) and a somewhat larger spot size (~30%
increase). Overall, our study establishes an experimental perspective on the
trade-off between a single grating normal incidence monochromator design and
the use of filters. As such, it provides a basis for selecting the most
appropriate approach in various fields where an easy-to-implement harmonic
selection from high harmonic generation is needed.
| arxiv topic:physics.optics |
arxiv_dataset-172892210.07545 | Hypergraphs for multiscale cycles in structured data
math.AT cs.CG q-bio.QM
Scientific data has been growing in both size and complexity across the
modern physical, engineering, life and social sciences. Spatial structure, for
example, is a hallmark of many of the most important real-world complex
systems, but its analysis is fraught with statistical challenges. Topological
data analysis can provide a powerful computational window on complex systems.
Here we present a framework to extend and interpret persistent homology
summaries to analyse spatial data across multiple scales. We introduce
hyperTDA, a topological pipeline that unifies local (e.g. geodesic) and global
(e.g. Euclidean) metrics without losing spatial information, even in the
presence of noise. Homology generators offer an elegant and flexible
description of spatial structures and can capture the information computed by
persistent homology in an interpretable way. Here the information computed by
persistent homology is transformed into a weighted hypergraph, where hyperedges
correspond to homology generators. We consider different choices of generators
(e.g. matroid or minimal) and find that centrality and community detection are
robust to either choice. We compare hyperTDA to existing geometric measures and
validate its robustness to noise. We demonstrate the power of computing
higher-order topological structures on spatial curves arising frequently in
ecology, biophysics, and biology, but also in high-dimensional financial
datasets. We find that hyperTDA can select between synthetic trajectories from
the landmark 2020 AnDi challenge and quantifies movements of different animal
species, even when data is limited.
| arxiv topic:math.AT cs.CG q-bio.QM |
arxiv_dataset-172902210.07645 | Controllable giant magneto resistance and perfect spin filtering in
$\alpha^{\prime}$-borophene nanoribbons
cond-mat.mes-hall
By using non-equilibrium Green's function (NEGF) method and tight-binding
(TB) approximation, we investigated a perfect control on spin transport in a
zigzag $\alpha^{\prime}$-boron nanoribbon ($\alpha^{\prime}$-BNR) as the must
semi-conducting structure of borophene. It has been found that when an
$\alpha^{\prime}$-BNR is exposed to an out-of-plane exchange magnetic field,
spin splitting occurs for both spin-up and spin-down states in specific ranges
of energy. Therefore, the spin polarization of current could be controlled by
adjusting the energy of incoming electrons by means of an external back-gate
voltage. We focus on the edge manipulation of $\alpha^{\prime}$-BNR by
ferro-magnetic (FM) or anti-ferromagnetic (AFM) exchange field which leads to
the emergence of a giant magneto resistance and a perfect spin filtering. Local
current provides the best picture of spin distribution of current in the
nanoribbon. In order to observe the response of the system to the proximity
effect of magnetic strips, we calculate the magnetic moment of each site. Then,
we show that applying a transverse or perpendicular electric field in the
presence of the exchange magnetic field gives another controlling tool on the
spin polarization of current in a constant energy. Finally, by simultaneous
effect of in-plane and out-of-plane exchange magnetic field on the edges of
nanoribbon, we reach a control on spin rotation in the scattering region. Our
investigation guarantees the $\alpha^{\prime}$-BNR as a promising two
dimensional (2D) structure for spintronic purposes.
| arxiv topic:cond-mat.mes-hall |
arxiv_dataset-172912210.07745 | Confidence estimation of classification based on the distribution of the
neural network output layer
cs.CL
One of the most common problems preventing the application of prediction
models in the real world is lack of generalization: The accuracy of models,
measured in the benchmark does repeat itself on future data, e.g. in the
settings of real business. There is relatively little methods exist that
estimate the confidence of prediction models. In this paper, we propose novel
methods that, given a neural network classification model, estimate uncertainty
of particular predictions generated by this model. Furthermore, we propose a
method that, given a model and a confidence level, calculates a threshold that
separates prediction generated by this model into two subsets, one of them
meets the given confidence level. In contrast to other methods, the proposed
methods do not require any changes on existing neural networks, because they
simply build on the output logit layer of a common neural network. In
particular, the methods infer the confidence of a particular prediction based
on the distribution of the logit values corresponding to this prediction. The
proposed methods constitute a tool that is recommended for filtering
predictions in the process of knowledge extraction, e.g. based on web
scrapping, where predictions subsets are identified that maximize the precision
on cost of the recall, which is less important due to the availability of data.
The method has been tested on different tasks including relation extraction,
named entity recognition and image classification to show the significant
increase of accuracy achieved.
| arxiv topic:cs.CL |
arxiv_dataset-172922210.07845 | Flame-state monitoring based on very low number of visible or infrared
images via few-shot learning
cs.CV physics.app-ph
The current success of machine learning on image-based combustion monitoring
is based on massive data, which is costly even impossible for industrial
applications. To address this conflict, we introduce few-shot learning in order
to achieve combustion monitoring and classification for the first time. Two
algorithms, Siamese Network coupled with k Nearest Neighbors (SN-kNN) and
Prototypical Network (PN), were tested. Rather than utilizing solely visible
images as discussed in previous studies, we also used Infrared (IR) images. We
analyzed the training process, test performance and inference speed of two
algorithms on both image formats, and also used t-SNE to visualize learned
features. The results demonstrated that both SN-kNN and PN were capable to
distinguish flame states from learning with merely 20 images per flame state.
The worst performance, which was realized by PN on IR images, still possessed
precision, accuracy, recall, and F1-score above 0.95. We showed that visible
images demonstrated more substantial differences between classes and presented
more consistent patterns inside the class, which made the training speed and
model performance better compared to IR images. In contrast, the relatively low
quality of IR images made it difficult for PN to extract distinguishable
prototypes, which caused relatively weak performance. With the entrire training
set supporting classification, SN-kNN performed well with IR images. On the
other hand, benefitting from the architecture design, PN has a much faster
speed in training and inference than SN-kNN. The presented work analyzed the
characteristics of both algorithms and image formats for the first time, thus
providing guidance for their future utilization in combustion monitoring tasks.
| arxiv topic:cs.CV physics.app-ph |
arxiv_dataset-172932210.07945 | A dense mini-Neptune orbiting the bright young star HD 18599
astro-ph.EP
Very little is known about the young planet population because the detection
of small planets orbiting young stars is obscured by the effects of stellar
activity and fast rotation which mask planets within radial velocity and
transit data sets. The few planets that have been discovered in young clusters
generally orbit stars too faint for any detailed follow-up analysis. Here we
present the characterization of a new mini-Neptune planet orbiting the bright
(V=9) and nearby K2 dwarf star, HD 18599. The planet candidate was originally
detected in TESS light curves from Sectors 2, 3, 29, and 30, with an orbital
period of 4.138~days. We then used HARPS and FEROS radial velocities, to find
the companion mass to be 25.5$\pm$4.6~M$_\oplus$. When we combine this with the
measured radius from TESS, of 2.70$\pm$0.05~R$_\oplus$, we find a high
planetary density of 7.1$\pm$1.4~g cm$^{-3}$. The planet exists on the edge of
the Neptune Desert and is the first young planet (300 Myr) of its type to
inhabit this region. Structure models argue for a bulk composition to consist
of 23% H$_2$O and 77% Rock and Iron. Future follow-up with large ground- and
space-based telescopes can enable us to begin to understand in detail the
characteristics of young Neptunes in the galaxy.
| arxiv topic:astro-ph.EP |
arxiv_dataset-172942210.08045 | Flux eruption events drive angular momentum transport in magnetically
arrested accretion flows
astro-ph.HE astro-ph.GA
We evolve two high-resolution general relativistic magnetohydrodynamic
(GRMHD) simulations of advection-dominated accretion flows around non-spinning
black holes (BHs), each over a duration $\sim 3\times 10^5\,GM_{\rm BH}/c^3$.
One model captures the evolution of a weakly magnetized (SANE) disk and the
other a magnetically arrested disk (MAD). Magnetic flux eruptions in the MAD
model push out gas from the disk and launch strong winds with outflow
efficiencies at times reaching $10\%$ of the incoming accretion power. Despite
the substantial power in these winds, average mass outflow rates remain small
out to a radius $\sim100\,GM_{\rm BH}/c^2$, only reaching $\sim 60-80\%$ of the
horizon accretion rate. The average outward angular momentum transport is
primarily radial in both modes of accretion, but with a clear distinction:
magnetic flux eruption-driven disk winds cause a strong vertical flow of
angular momentum in the MAD model, while for the SANE model, the
magnetorotational instability (MRI) moves angular momentum mostly equatorially
through the disk. Further, we find that the MAD state is highly transitory and
non-axisymmetric, with the accretion mode often changing to a SANE-like state
following an eruption before reattaining magnetic flux saturation with time.
The Reynolds stress changes direction during such transitions, with the MAD
(SANE) state showing an inward (outward) stress, possibly pointing to
intermittent MRI-driven accretion in MADs. Pinning down the nature of flux
eruptions using next-generation telescopes will be crucial in understanding the
flow of mass, magnetic flux and angular momentum in sub-Eddington accreting BHs
like M87$^*$ and Sagittarius A$^*$.
| arxiv topic:astro-ph.HE astro-ph.GA |
arxiv_dataset-172952210.08145 | Self-Repetition in Abstractive Neural Summarizers
cs.CL
We provide a quantitative and qualitative analysis of self-repetition in the
output of neural summarizers. We measure self-repetition as the number of
n-grams of length four or longer that appear in multiple outputs of the same
system. We analyze the behavior of three popular architectures (BART, T5, and
Pegasus), fine-tuned on five datasets. In a regression analysis, we find that
the three architectures have different propensities for repeating content
across output summaries for inputs, with BART being particularly prone to
self-repetition. Fine-tuning on more abstractive data, and on data featuring
formulaic language, is associated with a higher rate of self-repetition. In
qualitative analysis we find systems produce artefacts such as ads and
disclaimers unrelated to the content being summarized, as well as formulaic
phrases common in the fine-tuning domain. Our approach to corpus-level analysis
of self-repetition may help practitioners clean up training data for
summarizers and ultimately support methods for minimizing the amount of
self-repetition.
| arxiv topic:cs.CL |
arxiv_dataset-172962210.08245 | Medial and isospectral algebras
math.RA math.AC
The purpose of this paper is to give a systematic study of two new classes of
commutative nonassociative algebras, the so-called isospectral and medial
algebras. An isospectral algebra $\mathbb{A}$ is a generic commutative
nonassociative algebra whose idempotents have the same Peirce spectrum. A
medial algebra is algebra with identity $(xy)(zw)=(xz)(yw)$. We show that these
two classes are essentially coincide. We also prove that any medial spectral
algebra is isomorphic to a certain isotopic deformation of the commutative
associative quotient algebra $\mathbf{K}[z]/(z^n-1)$.
| arxiv topic:math.RA math.AC |
arxiv_dataset-172972210.08345 | Augmentation-Free Graph Contrastive Learning of Invariant-Discriminative
Representations
cs.LG cs.AI
The pretasks are mainly built on mutual information estimation, which
requires data augmentation to construct positive samples with similar semantics
to learn invariant signals and negative samples with dissimilar semantics in
order to empower representation discriminability. However, an appropriate data
augmentation configuration depends heavily on lots of empirical trials such as
choosing the compositions of data augmentation techniques and the corresponding
hyperparameter settings. We propose an augmentation-free graph contrastive
learning method, invariant-discriminative graph contrastive learning (iGCL),
that does not intrinsically require negative samples. iGCL designs the
invariant-discriminative loss (ID loss) to learn invariant and discriminative
representations. On the one hand, ID loss learns invariant signals by directly
minimizing the mean square error between the target samples and positive
samples in the representation space. On the other hand, ID loss ensures that
the representations are discriminative by an orthonormal constraint forcing the
different dimensions of representations to be independent of each other. This
prevents representations from collapsing to a point or subspace. Our
theoretical analysis explains the effectiveness of ID loss from the
perspectives of the redundancy reduction criterion, canonical correlation
analysis, and information bottleneck principle. The experimental results
demonstrate that iGCL outperforms all baselines on 5 node classification
benchmark datasets. iGCL also shows superior performance for different label
ratios and is capable of resisting graph attacks, which indicates that iGCL has
excellent generalization and robustness. The source code is available at
https://github.com/lehaifeng/T-GCN/tree/master/iGCL.
| arxiv topic:cs.LG cs.AI |
arxiv_dataset-172982210.08445 | Connection-Based Scheduling for Real-Time Intersection Control
cs.AI
We introduce a heuristic scheduling algorithm for real-time adaptive traffic
signal control to reduce traffic congestion. This algorithm adopts a lane-based
model that estimates the arrival time of all vehicles approaching an
intersection through different lanes, and then computes a schedule (i.e., a
signal timing plan) that minimizes the cumulative delay incurred by all
approaching vehicles. State space, pruning checks and an admissible heuristic
for A* search are described and shown to be capable of generating an
intersection schedule in real-time (i.e., every second). Due to the
effectiveness of the heuristics, the proposed approach outperforms a less
expressive Dynamic Programming approach and previous A*-based approaches in
run-time performance, both in simulated test environments and actual field
tests.
| arxiv topic:cs.AI |
arxiv_dataset-172992210.08545 | D/M/1 Queue: Policies and Control
math.PR cs.DM math.HO
Equilibrium G/M/1-FIFO waiting times are exponentially distributed, as first
proved by Smith (1953). For other client-sorting policies, such generality is
not feasible. Assume that interarrival times are constant. Symbolics for the
D/M/1-LIFO density are completely known; numerics for D/M/1-SIRO arise via an
unpublished recursion due to Burke (1967). Consider a weighted sum of two
costs, one from keeping clients waiting for treatment and the other from having
the server idle. With this in mind, what is the optimal interarrival time and
how does this depend on the choice of policy?
| arxiv topic:math.PR cs.DM math.HO |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.