id stringlengths 16 29 | text stringlengths 86 3.49k | source stringlengths 14 112 |
|---|---|---|
arxiv_dataset-119001910.06081 | Understanding and Pushing the Limits of the Elo Rating Algorithm
math.ST stat.TH
This work is concerned with the rating of players/teams in face-to-face games
with three possible outcomes: loss, win, and draw. This is one of the
fundamental problems in sport analytics, where the very simple and popular,
non-trivial algorithm was proposed by Arpad Elo in late fifties to rate chess
players. In this work we explain the mathematical model underlying the Elo
algorithm and, in particular, we explain what is the implicit but not yet
spelled out, assumption about the model of draws. We further extend the model
to provide flexibility and remove the unrealistic implicit assumptions of the
Elo algorithm. This yields the new rating algorithm, we call $\kappa$-Elo,
which is equally simple as the Elo algorithm but provides a possibility to
adjust to the frequency of draws. The discussion of the importance of the
appropriate choice of the parameters is carried out and illustrated using
results from English Premier League football seasons.
| arxiv topic:math.ST stat.TH |
arxiv_dataset-119011910.06181 | The critical role of the interaction potential and simulation protocol
for the structural and mechanical properties of sodosilicate glasses
cond-mat.dis-nn cond-mat.mtrl-sci
We compare the ability of various interaction potentials to predict the
structural and mechanical properties of silica and sodium silicate glasses.
While most structural quantities show a relatively mild dependence on the
potential used, the mechanical properties such as the failure stress and strain
as well as the elastic moduli depend very strongly on the potential, once
finite size effects have been taken into account. We find that to avoid such
finite size effects, samples of at least 75,000 atoms are needed. Finally we
probe how the simulation ensemble influences the fracture properties of the
glasses and conclude that fracture simulations should be carried out in the
constant pressure ensemble.
| arxiv topic:cond-mat.dis-nn cond-mat.mtrl-sci |
arxiv_dataset-119021910.06281 | Dynamic Complexity Meets Parameterised Algorithms
cs.LO cs.CC
Dynamic Complexity studies the maintainability of queries with logical
formulas in a setting where the underlying structure or database changes over
time. Most often, these formulas are from first-order logic, giving rise to the
dynamic complexity class DynFO. This paper investigates extensions of DynFO in
the spirit of parameterised algorithms. In this setting structures come with a
parameter $k$ and the extensions allow additional "space" of size $f(k)$ (in
the form of an additional structure of this size) or additional time $f(k)$ (in
the form of iterations of formulas) or both. The resulting classes are compared
with their non-dynamic counterparts and other classes. The main part of the
paper explores the applicability of methods for parameterised algorithms to
this setting through case studies for various well-known parameterised
problems.
| arxiv topic:cs.LO cs.CC |
arxiv_dataset-119031910.06381 | Principled estimation of regression discontinuity designs
stat.AP econ.EM
Regression discontinuity designs are frequently used to estimate the causal
effect of election outcomes and policy interventions. In these contexts,
treatment effects are typically estimated with covariates included to improve
efficiency. While including covariates improves precision asymptotically, in
practice, treatment effects are estimated with a small number of observations,
resulting in considerable fluctuations in treatment effect magnitude and
precision depending upon the covariates chosen. This practice thus incentivizes
researchers to select covariates which maximize treatment effect statistical
significance rather than precision. Here, I propose a principled approach for
estimating RDDs which provides a means of improving precision with covariates
while minimizing adverse incentives. This is accomplished by integrating the
adaptive LASSO, a machine learning method, into RDD estimation using an R
package developed for this purpose, adaptiveRDD. Using simulations, I show that
this method significantly improves treatment effect precision, particularly
when estimating treatment effects with fewer than 200 observations.
| arxiv topic:stat.AP econ.EM |
arxiv_dataset-119041910.06481 | Solitary wave solutions to the Isobe-Kakinuma model for water waves
math.AP
We consider the Isobe-Kakinuma model for two-dimensional water waves in the
case of the flat bottom. The Isobe-Kakinuma model is a system of Euler-Lagrange
equations for a Lagrangian approximating Luke's Lagrangian for water waves. We
show theoretically the existence of a family of small amplitude solitary wave
solutions to the Isobe-Kakinuma model in the long wave regime. Numerical
analysis for large amplitude solitary wave solutions is also provided and
suggests the existence of a solitary wave of extreme form with a sharp crest.
| arxiv topic:math.AP |
arxiv_dataset-119051910.06581 | Effects of coherence on quantum speed limits and shortcuts to
adiabaticity in many-particle systems
quant-ph cond-mat.quant-gas
We discuss the effects of many-body coherence on the speed of evolution of
ultracold atomic gases and the relation to quantum speed limits. Our approach
is focused on two related systems, spinless fermions and the bosonic
Tonks-Girardeau gas, which possess equivalent density dynamics but very
different coherence properties. To illustrate the effect of the coherence on
the dynamics we consider squeezing an anharmonic potential which confines the
particles and find that the speed of the evolution exhibits subtle, but
fundamental differences between the two systems. Furthermore, we explore the
difference in the driven dynamics by implementing a shortcut to adiabaticity
designed to reduce spurious excitations. We show that collisions between the
strongly interacting bosons can lead to changes in the coherence which results
in different evolution speeds and therefore different fidelities of the final
states.
| arxiv topic:quant-ph cond-mat.quant-gas |
arxiv_dataset-119061910.06681 | On the minimality of Keplerian arcs with fixed negative energy
math.CA
We revisit a classical result by Jacobi on the local minimality, as critical
points of the corresponding energy functional, of fixed-energy solutions of the
Kepler equation joining two distinct points with the same distance from the
origin. Our proof relies on the Morse index theorem, together with a
characterization of the conjugate points as points of geodesic bifurcation.
| arxiv topic:math.CA |
arxiv_dataset-119071910.06781 | Optimal principal component Analysis of STEM XEDS spectrum images
eess.IV physics.data-an
STEM XEDS spectrum images can be drastically denoised by application of the
principal component analysis (PCA). This paper looks inside the PCA workflow
step by step on an example of a complex semiconductor structure consisting of a
number of different phases. Typical problems distorting the principal
components decomposition are highlighted and solutions for the successful PCA
are described. Particular attention is paid to the optimal truncation of
principal components in the course of reconstructing denoised data. A novel
accurate and robust method, which overperforms the existing truncation methods
is suggested for the first time and described in details.
| arxiv topic:eess.IV physics.data-an |
arxiv_dataset-119081910.06881 | New bounds for the ratio of power means
math.CA
We show that for real numbers $p,\,q$ with $q<p$, and the related power means
$\mathscr{P}_p$, $\mathscr{P}_q$, the inequality
$$\frac{\mathscr{P}_p(x)}{\mathscr{P}_q(x)} \le \exp \bigg( \frac{p-q}8 \cdot
\bigg(\ln\bigg(\frac{\max x}{\min x}\bigg)\bigg)^2 \:\bigg)$$ holds for every
vector $x$ of positive reals. Moreover we prove that, for all such pairs
$(p,q)$, the constant $\tfrac{p-q}8$ is sharp.
| arxiv topic:math.CA |
arxiv_dataset-119091910.06981 | Design and Characterization of Q-enhanced Silicon Nitride Racetrack
Micro-Resonators
physics.optics physics.app-ph
Q-enhanced racetrack micro-resonators for the silicon nitride photonics
integration platform have been designed, fabricated and characterized. The
proposed geometries permit to mitigate the impact of radiation loss at curved
waveguides, one of the major limitations of silicon nitride circuits, therefore
providing an increase of the intrinsic Q factor of micro-resonators when
compared with the conventional structures with the same bent radii. The schemes
put forward in this work permit a reduction of the size of the devices that has
a direct impact on the integration scale in this platform. When used in the
curved sections of waveguides routing optical signals within an integrated
photonic circuit, these geometries provide a reduction of the radiation loss
and permit the use of smaller bent radii and to increase the circuit
integration density.
| arxiv topic:physics.optics physics.app-ph |
arxiv_dataset-119101910.07081 | Geometric Inequalities for Quasi-Local Masses
math.DG gr-qc
In this paper lower bounds are obtained for quasi-local masses in terms of
charge, angular momentum, and horizon area. In particular we treat three
quasi-local masses based on a Hamiltonian approach, namely the Brown-York,
Liu-Yau, and Wang-Yau masses. The geometric inequalities are motivated by
analogous results for the ADM mass. They may be interpreted as localized
versions of these inequalities, and are also closely tied to the conjectured
Bekenstein bounds for entropy of macroscopic bodies. In addition, we give a new
proof of the positivity property for the Wang-Yau mass which is used to remove
the spin condition in higher dimensions. Furthermore, we generalize a recent
result of Lu and Miao to obtain a localized version of the Penrose inequality
for the static Wang-Yau mass.
| arxiv topic:math.DG gr-qc |
arxiv_dataset-119111910.07181 | BERTRAM: Improved Word Embeddings Have Big Impact on Contextualized
Model Performance
cs.CL
Pretraining deep language models has led to large performance gains in NLP.
Despite this success, Schick and Sch\"utze (2020) recently showed that these
models struggle to understand rare words. For static word embeddings, this
problem has been addressed by separately learning representations for rare
words. In this work, we transfer this idea to pretrained language models: We
introduce BERTRAM, a powerful architecture based on BERT that is capable of
inferring high-quality embeddings for rare words that are suitable as input
representations for deep language models. This is achieved by enabling the
surface form and contexts of a word to interact with each other in a deep
architecture. Integrating BERTRAM into BERT leads to large performance
increases due to improved representations of rare and medium frequency words on
both a rare word probing task and three downstream tasks.
| arxiv topic:cs.CL |
arxiv_dataset-119121910.07281 | The diameter and radius of radially maximal graphs
math.CO
A graph is called radially maximal if it is not complete and the addition of
any new edge decreases its radius. In 1976 Harary and Thomassen proved that the
radius $r$ and diameter $d$ of any radially maximal graph satisfy $r\le d\le
2r-2.$ Dutton, Medidi and Brigham rediscovered this result with a different
proof in 1995 and they posed the conjecture that the converse is true, that is,
if $r$ and $d$ are positive integers satisfying $r\le d\le 2r-2,$ then there
exists a radially maximal graph with radius $r$ and diameter $d.$ We prove this
conjecture and a little more.
| arxiv topic:math.CO |
arxiv_dataset-119131910.07381 | Using learning analytics to provide personalized recommendations for
finding peers
cs.HC cs.CY
This work aims to propose a method to support students in finding appropriate
peers in collaborative and blended learning settings. The main goal of this
research is to bridge the gap between pedagogical theory and data driven
practice to provide personalized and adaptive guidance to students who engage
in computer supported learning activities. The research hypothesis is that we
can use Learning Analytics to model students' cognitive state and to assess
whether the student is in the Zone of Proximal Development. Based on this
assessment, we can plan how to provide scaffolding based on the principles of
Contingent Tutoring and how to form study groups based on the principles of the
Zone of Proximal Development.
| arxiv topic:cs.HC cs.CY |
arxiv_dataset-119141910.07481 | Using Whole Document Context in Neural Machine Translation
cs.CL
In Machine Translation, considering the document as a whole can help to
resolve ambiguities and inconsistencies. In this paper, we propose a simple yet
promising approach to add contextual information in Neural Machine Translation.
We present a method to add source context that capture the whole document with
accurate boundaries, taking every word into account. We provide this additional
information to a Transformer model and study the impact of our method on three
language pairs. The proposed approach obtains promising results in the
English-German, English-French and French-English document-level translation
tasks. We observe interesting cross-sentential behaviors where the model learns
to use document-level information to improve translation coherence.
| arxiv topic:cs.CL |
arxiv_dataset-119151910.07581 | Scaling up Psychology via Scientific Regret Minimization: A Case Study
in Moral Decisions
cs.CY cs.AI cs.LG
Do large datasets provide value to psychologists? Without a systematic
methodology for working with such datasets, there is a valid concern that
analyses will produce noise artifacts rather than true effects. In this paper,
we offer a way to enable researchers to systematically build models and
identify novel phenomena in large datasets. One traditional approach is to
analyze the residuals of models---the biggest errors they make in predicting
the data---to discover what might be missing from those models. However, once a
dataset is sufficiently large, machine learning algorithms approximate the true
underlying function better than the data, suggesting instead that the
predictions of these data-driven models should be used to guide model-building.
We call this approach "Scientific Regret Minimization" (SRM) as it focuses on
minimizing errors for cases that we know should have been predictable. We
demonstrate this methodology on a subset of the Moral Machine dataset, a public
collection of roughly forty million moral decisions. Using SRM, we found that
incorporating a set of deontological principles that capture dimensions along
which groups of agents can vary (e.g. sex and age) improves a computational
model of human moral judgment. Furthermore, we were able to identify and
independently validate three interesting moral phenomena: criminal
dehumanization, age of responsibility, and asymmetric notions of
responsibility.
| arxiv topic:cs.CY cs.AI cs.LG |
arxiv_dataset-119161910.07681 | Primary Vertex Selection in VBF Higgs to Invisibles at the HL-LHC with
the ATLAS Experiment
hep-ex
ATLAS has developed a new approach for vertex selection in VBF invisible
events at HL-LHC conditions, exploiting its new forward tracking capabilities,
integrating calorimeter and tracking information to mitigate the impact of
pileup vertex merging, and introducing a new way to apply pile-up jet
suppression methods for the selection of VBF jets. The new algorithm is
insensitive to pileup density and improves the average vertex selection
efficiency from 86% to 95% under tight VBF event selection cuts.
| arxiv topic:hep-ex |
arxiv_dataset-119171910.07781 | Econometric Models of Network Formation
econ.EM
This article provides a selective review on the recent literature on
econometric models of network formation. The survey starts with a brief
exposition on basic concepts and tools for the statistical description of
networks. I then offer a review of dyadic models, focussing on statistical
models on pairs of nodes and describe several developments of interest to the
econometrics literature. The article also presents a discussion of non-dyadic
models where link formation might be influenced by the presence or absence of
additional links, which themselves are subject to similar influences. This is
related to the statistical literature on conditionally specified models and the
econometrics of game theoretical models. I close with a (non-exhaustive)
discussion of potential areas for further development.
| arxiv topic:econ.EM |
arxiv_dataset-119181910.07881 | Improving Heart Rate Estimation on Consumer Grade Wrist-Worn Device
Using Post-Calibration Approach
eess.SP
The technological advancement in wireless health monitoring through the
direct contact of the skin allows the development of light-weight wrist-worn
wearable devices to be equipped with different sensors such as
photoplethysmography (PPG) sensors. However, the motion artifact (MA) is
possible to occur during daily activities. In this study, we attempted to
perform a post-calibration of the heart rate (HR) estimation during the three
possible states of average daily activity (resting, \textcolor{red}{laying
down}, and intense treadmill activity states) in 29 participants (130
minutes/person) on four popular wearable devices: Fitbit Charge HR, Apple Watch
Series 4, TicWatch Pro, and Empatica E4. In comparison to the standard
measurement (HR$_\text{ECG}$), HR provided by Fitbit Charge HR
(HR$_\text{Fitbit}$) yielded the highest error of $3.26 \pm 0.34$ bpm in
resting, $2.33 \pm 0.23$ bpm in \textcolor{red}{laying down}, $9.53 \pm 1.47$
bpm in intense treadmill activity states, and $5.02 \pm 0.64$ bpm in all states
combined among the four chosen devices. Following our improving HR estimation
model with rolling windows as feature (HR$_\text{R}$), the mean absolute error
(MAE) was significantly reduced by $33.44\%$ in resting, $15.88\%$ in
\textcolor{red}{laying down}, $9.55\%$ in intense treadmill activity states,
and $18.73\%$ in all states combined. This demonstrates the feasibility of our
proposed methods in order to correct and provide HR monitoring post-calibrated
with high accuracy, raising further awareness of individual fitness in the
daily application.
| arxiv topic:eess.SP |
arxiv_dataset-119191910.07981 | Mass bound for primordial black hole from trans-Planckian censorship
conjecture
astro-ph.CO gr-qc hep-th
The recently proposed trans-Planckian censorship conjecture (TCC) imposes a
strong constraint on the inflationary Hubble scale, of which the upper bound
could be largely relaxed by considering a noninstantaneous reheating history.
In this paper we will show that, if the primordial black holes (PBHs) are
formed at reentry in the radiation-dominated era from the enhanced curvature
perturbations at small scales, the TCC would impose a lower bound on the PBH
mass $M_\mathrm{PBH}>\gamma(H_\mathrm{end}/10^9\,\mathrm{GeV})^2\,M_\odot$
regardless of the details for reheating history, where $\gamma$ is the collapse
efficiency factor and $H_\mathrm{end}$ is the Hubble scale at the end of
inflation. In particular, the current open window for PBHs to make up all the
cold dark matter could be totally ruled out if the inflationary Hubble scale is
larger than 10 TeV. For the case of PBHs formed in an early matter-dominated
era, an upper mass bound is obtained.
| arxiv topic:astro-ph.CO gr-qc hep-th |
arxiv_dataset-119201910.08081 | Many-body effects in nodal-line semimetals: correction to the optical
conductivity
cond-mat.str-el cond-mat.mes-hall hep-th
Coulomb interaction might have important effects on the physical observables
in topological semimetals with vanishing density of states at the band touching
due to the weak screening. In this work, we show that Kohn's theorem is not
fulfilled in nodal-line semimetals (NLSMs), which implies non-vanishing
interaction corrections to the conductivity. Using renormalized perturbation
theory, we determine the first-order optical conductivity in a clean NLSM to be
$\sigma_{\perp \perp}(\Omega) = 2 \sigma_{\parallel \parallel}(\Omega) =
\sigma_0 [1 + C_2 \alpha_R(\Omega)]$, where $\perp$ and $\parallel$ denote the
perpendicular and parallel components with respect to the nodal loop, $\sigma_0
= (2 \pi k_0) e^2/(16h)$ is the conductivity in the noninteracting limit, $2
\pi k_0$ is the nodal loop perimeter, $C_2 = (19-6\pi)/12 \simeq 0.013$ is a
numerical constant and $\alpha_R(\Omega)$ is the renormalized fine structure
constant in the NLSM. The analogies between NLSMs and 2D Dirac fermions are
reflected in the universal character of the correction $C_2 \alpha_R(\Omega)$,
which is exactly parallel to that of graphene. Finally, we analyze some
experiments that have determined the optical conductivity in NLSMs, discussing
the possibility of experimentally measuring our result.
| arxiv topic:cond-mat.str-el cond-mat.mes-hall hep-th |
arxiv_dataset-119211910.08181 | Online Learning in Planar Pushing with Combined Prediction Model
cs.RO cs.LG
Pushing is a useful robotic capability for positioning and reorienting
objects. The ability to accurately predict the effect of pushes can enable
efficient trajectory planning and complicated object manipulation. Physical
prediction models for planar pushing have long been established, but their
assumptions and requirements usually don't hold in most practical settings.
Data-driven approaches can provide accurate predictions for offline data, but
they often have generalizability issues. In this paper, we propose a combined
prediction model and an online learning framework for planar push prediction.
The combined model consists of a neural network module and analytical
components with a low-dimensional parameter. We train the neural network
offline using pre-collected pushing data. In online situations, the
low-dimensional analytical parameter is learned directly from online pushes to
quickly adapt to the new environments. We test our combined model and learning
framework on real pushing experiments. Our experimental results show that our
model is able to quickly adapt to new environments while achieving similar
final prediction performance as that of pure neural network models.
| arxiv topic:cs.RO cs.LG |
arxiv_dataset-119221910.08281 | Point Process Flows
cs.LG stat.ML
Event sequences can be modeled by temporal point processes (TPPs) to capture
their asynchronous and probabilistic nature. We propose an intensity-free
framework that directly models the point process distribution by utilizing
normalizing flows. This approach is capable of capturing highly complex
temporal distributions and does not rely on restrictive parametric forms.
Comparisons with state-of-the-art baseline models on both synthetic and
challenging real-life datasets show that the proposed framework is effective at
modeling the stochasticity of discrete event sequences.
| arxiv topic:cs.LG stat.ML |
arxiv_dataset-119231910.08381 | Model Compression with Two-stage Multi-teacher Knowledge Distillation
for Web Question Answering System
cs.CL
Deep pre-training and fine-tuning models (such as BERT and OpenAI GPT) have
demonstrated excellent results in question answering areas. However, due to the
sheer amount of model parameters, the inference speed of these models is very
slow. How to apply these complex models to real business scenarios becomes a
challenging but practical problem. Previous model compression methods usually
suffer from information loss during the model compression procedure, leading to
inferior models compared with the original one. To tackle this challenge, we
propose a Two-stage Multi-teacher Knowledge Distillation (TMKD for short)
method for web Question Answering system. We first develop a general Q\&A
distillation task for student model pre-training, and further fine-tune this
pre-trained student model with multi-teacher knowledge distillation on
downstream tasks (like Web Q\&A task, MNLI, SNLI, RTE tasks from GLUE), which
effectively reduces the overfitting bias in individual teacher models, and
transfers more general knowledge to the student model. The experiment results
show that our method can significantly outperform the baseline methods and even
achieve comparable results with the original teacher models, along with
substantial speedup of model inference.
| arxiv topic:cs.CL |
arxiv_dataset-119241910.08481 | A model problem for quasinormal ringdown on asymptotically flat or
extremal black holes
math.AP gr-qc
We consider a wave equation with a potential on the half-line as a model
problem for wave propagation close to an extremal horizon, or the
asymptotically flat end of a black hole spacetime. We propose a definition of
quasinormal frequencies (QNFs) as eigenvalues of the generator of time
translations for a null foliation, acting on an appropriate (Gevrey based)
Hilbert space. We show that this QNF spectrum is discrete in a subset of
$\mathbb{C}$ which includes the region $\{$Re$(s) >-b$, $|$Im $(s)|> K\}$ for
any $b>0$ and some $K=K(b) \gg 1$. As a corollary we establish the
meromorphicity of the scattering resolvent in a sector $|$arg$(s)| <\varphi_0$
for some $\varphi_0 > \frac{2\pi}{3}$, and show that the poles occur only at
quasinormal frequencies according to our definition. This result applies in
situations where the method of complex scaling cannot be directly applied, as
our potentials need not be analytic. Finally, we show that QNFs computed by the
continued fraction method of Leaver are necessarily QNFs according to our new
definition. This paper is a companion to [D. Gajic and C. Warnick, Quasinormal
modes in extremal Reissner-Nordstr\"om spacetimes, preprint (2019)], which
deals with the QNFs of the wave equation on the extremal Reissner-Nordstr\"om
black hole.
| arxiv topic:math.AP gr-qc |
arxiv_dataset-119251910.08581 | Towards Quantifying Intrinsic Generalization of Deep ReLU Networks
cs.LG cs.NE stat.ML
Understanding the underlying mechanisms that enable the empirical successes
of deep neural networks is essential for further improving their performance
and explaining such networks. Towards this goal, a specific question is how to
explain the "surprising" behavior of the same over-parametrized deep neural
networks that can generalize well on real datasets and at the same time
"memorize" training samples when the labels are randomized. In this paper, we
demonstrate that deep ReLU networks generalize from training samples to new
points via piece-wise linear interpolation. We provide a quantified analysis on
the generalization ability of a deep ReLU network: Given a fixed point
$\mathbf{x}$ and a fixed direction in the input space $\mathcal{S}$, there is
always a segment such that any point on the segment will be classified the same
as the fixed point $\mathbf{x}$. We call this segment the $generalization \
interval$. We show that the generalization intervals of a ReLU network behave
similarly along pairwise directions between samples of the same label in both
real and random cases on the MNIST and CIFAR-10 datasets. This result suggests
that the same interpolation mechanism is used in both cases. Additionally, for
datasets using real labels, such networks provide a good approximation of the
underlying manifold in the data, where the changes are much smaller along
tangent directions than along normal directions. On the other hand, however,
for datasets with random labels, generalization intervals along mid-lines of
triangles with the same label are much smaller than those on the datasets with
real labels, suggesting different behaviors along other directions. Our
systematic experiments demonstrate for the first time that such deep neural
networks generalize through the same interpolation and explain the differences
between their performance on datasets with real and random labels.
| arxiv topic:cs.LG cs.NE stat.ML |
arxiv_dataset-119261910.08681 | SPARK: Spatial-aware Online Incremental Attack Against Visual Tracking
cs.CV
Adversarial attacks of deep neural networks have been intensively studied on
image, audio, natural language, patch, and pixel classification tasks.
Nevertheless, as a typical, while important real-world application, the
adversarial attacks of online video object tracking that traces an object's
moving trajectory instead of its category are rarely explored. In this paper,
we identify a new task for the adversarial attack to visual tracking: online
generating imperceptible perturbations that mislead trackers along an incorrect
(Untargeted Attack, UA) or specified trajectory (Targeted Attack, TA). To this
end, we first propose a \textit{spatial-aware} basic attack by adapting
existing attack methods, i.e., FGSM, BIM, and C&W, and comprehensively analyze
the attacking performance. We identify that online object tracking poses two
new challenges: 1) it is difficult to generate imperceptible perturbations that
can transfer across frames, and 2) real-time trackers require the attack to
satisfy a certain level of efficiency. To address these challenges, we further
propose the spatial-aware online incremental attack (a.k.a. SPARK) that
performs spatial-temporal sparse incremental perturbations online and makes the
adversarial attack less perceptible. In addition, as an optimization-based
method, SPARK quickly converges to very small losses within several iterations
by considering historical incremental perturbations, making it much more
efficient than basic attacks. The in-depth evaluation on state-of-the-art
trackers (i.e., SiamRPN++ with AlexNet, MobileNetv2, and ResNet-50, and SiamDW)
on OTB100, VOT2018, UAV123, and LaSOT demonstrates the effectiveness and
transferability of SPARK in misleading the trackers under both UA and TA with
minor perturbations.
| arxiv topic:cs.CV |
arxiv_dataset-119271910.08781 | Development of a high-resolution and high efficiency Single Photon
detector for studying cardiovascular diseases in mice
physics.med-ph physics.ins-det
SPECT systems using pinhole apertures permit radiolabeled molecular
distributions to be imaged in vivo in small animals. Nevertheless studying
cardiovascular diseases by means of small animal models is very challenging.
Specifically, submillimeter spatial resolution, good energy resolution and high
sensitivity are required. We designed what we consider the "optimal"
radionuclide detector system for this task. It should allow studying both
detection of unstable atherosclerotic plaques and monitoring the effect of
therapies. Using mice is particularly challenging in situations that require
several intravenous injections of radiotracers, possibly for week or even
months, in chronically ill animals. Thus, alternative routes of delivering the
radiotracer in tail vein should be investigated. In this study we have
performed preliminary measurements of detection of atherosclerotic plaques in
genetically modified mice with high-resolution prototype detector. We have also
evaluated the feasibility of assessing left ventricular perfusion by
intraperitoneal delivering of MIBI-Tc in healthy mice.
| arxiv topic:physics.med-ph physics.ins-det |
arxiv_dataset-119281910.08881 | EQSA: Earthquake Situational Analytics from Social Media
cs.IR cs.SI
This paper introduces EQSA, an interactive exploratory tool for earthquake
situational analytics using social media. EQSA is designed to support users to
characterize the condition across the area around the earthquake zone,
regarding related events, resources to be allocated, and responses from the
community. On the general level, changes in the volume of messages from chosen
categories are presented, assisting users in conveying a general idea of the
condition. More in-depth analysis is provided with topic evolution, community
visualization, and location representation. EQSA is developed with intuitive,
interactive features and multiple linked views, visualizing social media data,
and supporting users to gain a comprehensive insight into the situation. In
this paper, we present the application of EQSA with the VAST Challenge 2019:
Mini-Challenge 3 (MC3) dataset.
| arxiv topic:cs.IR cs.SI |
arxiv_dataset-119291910.08981 | The prime number race and zeros of Dirichlet L-functions off the
critical line, II
math.NT
We continue our examination the effects of certain hypothetical
configurations of zeros of Dirichlet $L$-functions lying off the critical line
("barriers") on the relative magnitude of the functions $\pi_{q,a}(x)$. Here
$\pi_{q,a}(x)$ is the number of primes $\le x$ in the progression $a \mod q$.
In particular, we construct barriers so that $\pi_{q,1}(x)$ is simultaneously
greater than, or simultaneously less than, each of $k$ functions
$\pi_{q,a_i}(x)$ ($1\le i\le k$). We also construct barriers so that only a
small number of the $r!$ possible orderings of functions $\pi_{q,a_i}(x)$
($1\le i\le r$) occur for large $x$; see Theorem 5.1.
| arxiv topic:math.NT |
arxiv_dataset-119301910.09081 | An Application of Abel's Method to the Inverse Radon Transform
math.NA cs.NA
A method of approximating the inverse Radon transform on the plane by
integrating against a smooth kernel is investigated. For piecewise smooth
integrable functions, convergence theorems are proven and Gibbs phenomena are
ruled out. Geometric properties of the kernel and their implications for
computer implementation are discussed. Suggestions are made for applications
and an example is presented.
| arxiv topic:math.NA cs.NA |
arxiv_dataset-119311910.09181 | Alias Sampling Effect on the Calculation of MHD Mode Number in Fusion
Plasma
physics.plasm-ph
Detailed alias frequency formula and the effect of alias sampling on the
calculation of MHD mode number are derived. It is discovered that the absolute
MHD mode number/structure does not change under alias sampling. This discovery
can help us to determine the structure of the high frequency MHD mode with low
frequency sampled diagnostics even when the Nyquist-Shannon sampling theorem is
no longer valid.
| arxiv topic:physics.plasm-ph |
arxiv_dataset-119321910.09281 | Dealing with Sparse Rewards in Reinforcement Learning
cs.LG cs.AI stat.ML
Successfully navigating a complex environment to obtain a desired outcome is
a difficult task, that up to recently was believed to be capable only by
humans. This perception has been broken down over time, especially with the
introduction of deep reinforcement learning, which has greatly increased the
difficulty of tasks that can be automated. However, for traditional
reinforcement learning agents this requires an environment to be able to
provide frequent extrinsic rewards, which are not known or accessible for many
real-world environments. This project aims to explore and contrast existing
reinforcement learning solutions that circumnavigate the difficulties of an
environment that provide sparse rewards. Different reinforcement solutions will
be implemented over a several video game environments with varying difficulty
and varying frequency of rewards, as to properly investigate the applicability
of these solutions. This project introduces a novel reinforcement learning
solution by combining aspects of two existing state of the art sparse reward
solutions, curiosity driven exploration and unsupervised auxiliary tasks.
| arxiv topic:cs.LG cs.AI stat.ML |
arxiv_dataset-119331910.09381 | Superconductivity in Sn$_{1-x}$In$_{x}$Te thin films grown by molecular
beam epitaxy
cond-mat.mtrl-sci cond-mat.supr-con
The superconductor Sn$_{1-x}$In$_{x}$Te is derived from the topological
crystalline insulator SnTe and is a candidate topological superconductor. So
far, high-quality thin films of this material have not been available, even
though such samples would be useful for addressing the nature of its
superconductivity. Here we report the successful molecular beam epitaxy growth
of superconducting Sn$_{1-x}$In$_{x}$Te films by using Bi$_2$Te$_3$ as a buffer
layer. The data obtained from tunnel junctions made on such films show the
appearance of two superconducting gaps, which points to the coexistence of bulk
and surface superconductivity. Given the spin-momentum locking of the surface
states, the surface superconductivity is expected to be topological with an
effective $p$-wave character. Since the topological surface states of SnTe
consist of four Dirac cones, this platform offers an interesting playground for
studying topological surface superconductivity with additional degrees of
freedom.
| arxiv topic:cond-mat.mtrl-sci cond-mat.supr-con |
arxiv_dataset-119341910.09481 | Generation of exchange magnons in thin ferromagnetic films by ultrashort
acoustic pulses
cond-mat.mes-hall cond-mat.mtrl-sci cond-mat.other cond-mat.str-el
We investigate generation of exchange magnons by ultrashort, picosecond
acoustic pulses propagating through ferromagnetic thin films. Using the
Landau-Lifshitz-Gilbert equations we derive the dispersion relation for
exchange magnons for an external magnetic field tilted with respect to the film
normal. Decomposing the solution in a series of standing spin wave modes, we
derive a system of ordinary differential equations and driven harmonic
oscillator equations describing the dynamics of individual magnon mode. The
external magnetoelastic driving force is given by the time-dependent spatial
Fourier components of acoustic strain pulses inside the layer. Dependencies of
the magnon excitation efficiencies on the duration of the acoustic pulses and
the external magnetic field highlight the role of acoustic bandwidth and
phonon-magnon phase matching. Our simulations for ferromagnetic nickel evidence
the possibility of ultrafast magneto-acoustic excitation of exchange magnons
within the bandwidth of acoustic pulses in thin samples under conditions
readily obtained in femtosecond pump-probe experiments.
| arxiv topic:cond-mat.mes-hall cond-mat.mtrl-sci cond-mat.other cond-mat.str-el |
arxiv_dataset-119351910.09581 | Four-top as probe of light top-philic New Physics
hep-ph hep-ex
We study the four-top final state at the LHC as a probe for New Physics (NP)
effects due to new particles that couple predominantly to the top quark and
whose masses are below the top-quark-pair production threshold. We consider
simple NP models containing a new particle with either spin 0, spin 1, or spin
2, and find benchmark points compatible with current experimental results. We
find that interference effects between NP and QED amplitudes can be large,
pointing out the necessity of NLO contributions to be explicitly computed and
taken into account when NP is present. We examine kinematic differences between
these models and the Standard Model (SM) at the parton level and the
reconstructed level. In the latter case, we focus on events selected requiring
two same-sign leptons and multiple jets. We investigate how the different
Lorentz structure of the light NP affects the kinematic hardness, the
polarization, the spin correlations, and the angular distributions of the
parton-level and/or final-state particles. We find that spin-2 light NP would
be identified by harder kinematics than the SM. We also show that the angular
separation between the same-sign leptons is a sensitive observable for spin-0
NP. The spin-0 and spin-2 NP cases would also yield a signal in $t\bar t
\gamma\gamma$ with the invariant mass of the photons indicating the mass of the
new particle. The spin-1 NP would be identified through an excess in four-top
signal and slight or not modification in other observables, as for instance the
lack of signal in $t\bar t \gamma\gamma$ due to the Landau-Yang theorem. We
comment on the opportunities that would open from the kinematic reconstruction
of some of the top quarks in the four-top state. Our results provide new
handles to probe for light top-philic NP as part of the ongoing experimental
program of searches for four-top production at the LHC Run 2 and beyond.
| arxiv topic:hep-ph hep-ex |
arxiv_dataset-119361910.09681 | Effective energy density determines the dynamics of suspensions of
active and passive matter
physics.comp-ph cond-mat.soft q-bio.CB
The unique properties of suspensions containing both active (self-propelling)
and passive matter, arising from the nonequilibrium nature of these systems,
have been widely studied (e.g., enhanced diffusion, phase separation, and
directed motion). Despite this, our understanding of the specific roles played
by the relevant parameters of the constituent particles remains incomplete. For
instance, to what extent are the velocity and density of swimmers qualitatively
distinguishable when it comes to the resultant properties of the suspension as
a whole, and when are they merely two different realizations of the same thing?
Through the use of numerical simulations, containing both steric and
hydrodynamic interactions, we investigate a new parameter, the effective energy
density, and its ability to uniquely describe the dynamics and properties of a
hybrid system of active and passive particles, including the rate of pair
formation and the energy distribution amongst different constituent elements.
This parameter depends on both the density and the swimming velocity of the
active elements, unifying them into a single variable that surpasses the
descriptive ability of either alone.
| arxiv topic:physics.comp-ph cond-mat.soft q-bio.CB |
arxiv_dataset-119371910.09781 | An expression for the eddy field in a circular vacuum chamber for HEPS
booster dipole
physics.acc-ph
The analytical expression of the magnetic field distribution within the
aperture of a circular vacuum chamber due to the induced eddy is derived. Two
cases are discussed, one is the absence of iron, the other is that the vacuum
chamber is between the iron poles, that implies the use of the image current
methods. The current angular distribution in the vacuum chamber can be
calculated from the ramping rate of the exposed field, then the contour
integration is applied to the circular current to obtain the field
distribution. The formula can be used to estimate the undesired fields from a
circular beam box when it is exposed to a ramping field.
| arxiv topic:physics.acc-ph |
arxiv_dataset-119381910.09881 | Measuring the Hubble constant from the cooling of the CMB monopole
astro-ph.CO astro-ph.IM
The cosmic microwave background (CMB) monopole temperature evolves with the
inverse of the cosmological scale factor, independent of many cosmological
assumptions. With sufficient sensitivity, real-time cosmological observations
could thus be used to measure the local expansion rate of the Universe using
the cooling of the CMB. We forecast how well a CMB spectrometer could determine
the Hubble constant via this method. The primary challenge of such a mission
lies in the separation of Galactic and extra-Galactic foreground signals from
the CMB at extremely high precision. However, overcoming these obstacles could
potentially provide an independent, highly robust method to shed light on the
current low-/high-$z$ Hubble tension. An experiment with 3000 linearly spaced
bins between 5 GHz and 3 THz with a sensitivity of 1
$\mathrm{mJy\sqrt{yr}~sr^{-1}}$ per bin, could measure $H_0$ to 3% over a 10
year mission, given current foreground complexity. This sensitivity would also
enable high-precision measurements of the expected $\Lambda$CDM spectral
distortions, but remains futuristic at this stage.
| arxiv topic:astro-ph.CO astro-ph.IM |
arxiv_dataset-119391910.09981 | Two-particle contributions and nonlocal effects in the QCD sum rules for
the axialvector tetraquark candidate $Z_c(3900)$
hep-ph
In this article, we study the $Z_c(3900)$ with the QCD sum rules in details
by including the two-particle scattering state contributions and nonlocal
effects between the diquark and antidiquark constituents. The two-particle
scattering state contributions cannot saturate the QCD sum rules at the hadron
side, the contribution of the $Z_c(3900)$ plays an un-substitutable role, we
can saturate the QCD sum rules with or without the two-particle scattering
state contributions. If there exists a barrier or spatial separation between
the diquark and antidiquark constituents, the Feynman diagrams can be divided
into the factorizable and nonfactorizable diagrams. The factorizable diagrams
consist of two colored clusters and lead to a stable tetraquark state. The
nonfactorizable Feynman diagrams correspond to the tunnelling effects, which
play a minor important role in the QCD sum rules, and are consistent with the
small width of the $Z_c(3900)$. It is feasible to apply the QCD sum rules to
study the tetraquark states, which begin to receive contributions at the order
$\mathcal{O}(\alpha_s^0)$, not at the order $\mathcal{O}(\alpha_s^2)$.
| arxiv topic:hep-ph |
arxiv_dataset-119401910.10081 | The Evaluation of a Novel Asymptotic Solution to the Sommerfeld
Radiation Problem using an Efficient Method for the Calculation of Sommerfeld
Integrals in the Spectral Domain
eess.SP physics.comp-ph
In this work, a recently developed novel solution of the famous "Sommerfeld
Radiation Problem" is revisited. The solution is based on an analysis performed
entirely in the spectral domain, through which a compact asymptotic formula
describes the behavior of the EM field, which emanates from a vertical Hertzian
radiating dipole, located above flat, lossy ground. The paper is divided into
two parts. First, we demonstrate an efficient technique for the accurate
numeric calculation of the well - known Sommerfeld integrals, required for the
evaluation of the field. The results are compared against alternative
calculation approaches and validated with the corresponding Norton figures for
the Surface Wave. Then, in the second part, we briefly introduce the asymptotic
solution of interest and investigate its performance; we contrast the solution
versus the accurate numerical evaluation for the total received EM field and
also with a more basic asymptotic solution to the given problem, obtained via
the application of the Stationary Phase Method (SPM). Simulations for various
frequencies, distances, altitudes and ground characteristics are illustrated
and inferences for the applicability of the solution are made. Finally, special
cases, leading to analytic field expressions, close as well as far from the
interface, are examined.
| arxiv topic:eess.SP physics.comp-ph |
arxiv_dataset-119411910.10181 | Stable limit theorems on the Poisson space
math.PR
We prove limit theorems for functionals of a Poisson point process using the
Malliavin calculus on the Poisson space. The target distribution is
conditionally either a Gaussian vector or a Poisson random variable. The
convergence is stable and our conditions are expressed in terms of the
Malliavin operators. For conditionally Gaussian limits, we also obtain
quantitative bounds, given for the Monge-Kantorovich transport distance in the
univariate case; and for another probabilistic variational distance in higher
dimension. Our work generalizes several limit theorems on the Poisson space,
including the seminal works by Peccati, Sol\'e, Taqqu & Utzet for Gaussian
approximations; and by Peccati for Poisson approximations; as well as the
recently established fourth-moment theorem on the Poisson space of D\"obler &
Peccati. We give an application to stochastic processes.
| arxiv topic:math.PR |
arxiv_dataset-119421910.10281 | A Search-based Neural Model for Biomedical Nested and Overlapping Event
Detection
cs.CL
We tackle the nested and overlapping event detection task and propose a novel
search-based neural network (SBNN) structured prediction model that treats the
task as a search problem on a relation graph of trigger-argument structures.
Unlike existing structured prediction tasks such as dependency parsing, the
task targets to detect DAG structures, which constitute events, from the
relation graph. We define actions to construct events and use all the beams in
a beam search to detect all event structures that may be overlapping and
nested. The search process constructs events in a bottom-up manner while
modelling the global properties for nested and overlapping structures
simultaneously using neural networks. We show that the model achieves
performance comparable to the state-of-the-art model Turku Event Extraction
System (TEES) on the BioNLP Cancer Genetics (CG) Shared Task 2013 without the
use of any syntactic and hand-engineered features. Further analyses on the
development set show that our model is more computationally efficient while
yielding higher F1-score performance.
| arxiv topic:cs.CL |
arxiv_dataset-119431910.10381 | Proofs of Urysohn's Lemma and the Tietze Extension Theorem via the
Cantor function
math.GN
Urysohn's Lemma is a crucial property of normal spaces that deals with
separation of closed sets by continuous functions. It is also a fundamental
ingredient in proving the Tietze Extension Theorem, another property of normal
spaces that deals with the existence of extensions of continuous functions.
Using the Cantor function, we give alternative proofs for Urysohn's Lemma and
the Tietze Extension Theorem.
| arxiv topic:math.GN |
arxiv_dataset-119441910.10481 | The Task Analysis Cell Assembly Perspective
cs.HC cs.AI
An entirely novel synthesis combines the applied cognitive psychology of a
task analytic approach with a neural cell assembly perspective that models both
brain and mind function during task performance; similar cell assemblies could
be implemented as an artificially intelligent neural network. A simplified cell
assembly model is introduced and this leads to several new representational
formats that, in combination, are demonstrated as suitable for analysing tasks.
The advantages of using neural models are exposed and compared with previous
research that has used symbolic artificial intelligence production systems,
which make no attempt to model neurophysiology. For cognitive scientists, the
approach provides an easy and practical introduction to thinking about brains,
minds and artificial intelligence in terms of cell assemblies. In the future,
subsequent developments have the potential to lead to a new, general theory of
psychology and neurophysiology, supported by cell assembly based artificial
intelligences.
| arxiv topic:cs.HC cs.AI |
arxiv_dataset-119451910.10581 | A general approach to maximise information density in neutron
reflectometry analysis
cond-mat.soft physics.comp-ph stat.AP
Neutron and X-ray reflectometry are powerful techniques facilitating the
study of the structure of interfacial materials. The analysis of these
techniques is ill-posed in nature requiring the application of a
model-dependent methods. This can lead to the over- and under- analysis of
experimental data, when too many or too few parameters are allowed to vary in
the model. In this work, we outline a robust and generic framework for the
determination of the set of free parameters that is capable of maximising the
in-formation density of the model. This framework involves the determination of
the Bayesian evidence for each permutation of free parameters; and is applied
to a simple phospholipid monolayer. We believe this framework should become an
important component in reflectometry data analysis, and hope others more
regularly consider the relative evidence for their analytical models.
| arxiv topic:cond-mat.soft physics.comp-ph stat.AP |
arxiv_dataset-119461910.10681 | Continuous Control Set Nonlinear Model Predictive Control of Reluctance
Synchronous Machines
eess.SY cs.SY
In this paper we describe the design and implementation of a current
controller for a reluctance synchronous machine based on continuous set
nonlinear model predictive control. A computationally efficient grey box model
of the flux linkage map is employed in a tracking formulation which is
implemented using the high-performance framework for nonlinear model predictive
control acados. The resulting controller is validated in simulation and
deployed on a dSPACE real-time system connected to a physical reluctance
synchronous machine. Experimental results are presented where the proposed
implementation can reach sampling times in the range typical for electrical
drives and can achieve large improvements in terms of control performance with
respect to state-of-the-art classical control strategies.
| arxiv topic:eess.SY cs.SY |
arxiv_dataset-119471910.10781 | Hierarchical Transformers for Long Document Classification
cs.CL cs.LG stat.ML
BERT, which stands for Bidirectional Encoder Representations from
Transformers, is a recently introduced language representation model based upon
the transfer learning paradigm. We extend its fine-tuning procedure to address
one of its major limitations - applicability to inputs longer than a few
hundred words, such as transcripts of human call conversations. Our method is
conceptually simple. We segment the input into smaller chunks and feed each of
them into the base model. Then, we propagate each output through a single
recurrent layer, or another transformer, followed by a softmax activation. We
obtain the final classification decision after the last segment has been
consumed. We show that both BERT extensions are quick to fine-tune and converge
after as little as 1 epoch of training on a small, domain-specific data set. We
successfully apply them in three different tasks involving customer call
satisfaction prediction and topic classification, and obtain a significant
improvement over the baseline models in two of them.
| arxiv topic:cs.CL cs.LG stat.ML |
arxiv_dataset-119481910.10881 | Superposition as Data Augmentation using LSTM and HMM in Small Training
Sets
cs.LG eess.AS
Considering audio and image data as having quantum nature (data are
represented by density matrices), we achieved better results on training
architectures such as 3-layer stacked LSTM and HMM by mixing training samples
using superposition augmentation and compared with plain default training and
mix-up augmentation. This augmentation technique originates from the mix-up
approach but provides more solid theoretical reasoning based on quantum
properties. We achieved 3% improvement (from 68% to 71%) by using 38% lesser
number of training samples in Russian audio-digits recognition task and 7,16%
better accuracy than mix-up augmentation by training only 500 samples using HMM
on the same task. Also, we achieved 1.1% better accuracy than mix-up on first
900 samples in MNIST using 3-layer stacked LSTM.
| arxiv topic:cs.LG eess.AS |
arxiv_dataset-119491910.10981 | Recurrence coefficients for discrete orthogonal polynomials with
hypergeometric weight and discrete Painlev\'e equations
nlin.SI math-ph math.AG math.CA math.MP
Over the last decade it has become clear that discrete Painlev\'e equations
appear in a wide range of important mathematical and physical problems. Thus,
the question of recognizing a given non-autonomous recurrence as a discrete
Painlev\'e equation and determining its type according to Sakai's
classification scheme, understanding whether it is equivalent to some known
(model) example, and especially finding an explicit change of coordinates
transforming it to such an example, becomes one of the central ones.
Fortunately, Sakai's geometric theory provides an almost algorithmic procedure
for answering this question. In this paper we illustrate this procedure by
studying an example coming from the theory of discrete orthogonal polynomials.
There are many connections between orthogonal polynomials and Painlev\'e
equations, both differential and discrete. In particular, often the
coefficients of three-term recurrence relations for discrete orthogonal
polynomials can be expressed in terms of solutions of discrete Painlev\'e
equations. In this work we study discrete orthogonal polynomials with general
hypergeometric weight and show that their recurrence coefficients satisfy,
after some change of variables, the standard discrete Painlev\'e-V equation. We
also provide an explicit change of variables transforming this equation to the
standard form.
| arxiv topic:nlin.SI math-ph math.AG math.CA math.MP |
arxiv_dataset-119501910.11081 | Black Hole Horizons as Patternless Binary Messages and Markers of
Dimensionality
physics.gen-ph
This study aims to reconcile quantum theory with the universality of the
speed of light in vacuum and its implications on relativity through an
information-theoretic approach. We introduce the concepts of a holographic
sphere and variational potential. Entropy variation expressed in terms of the
information capacity of this sphere results in the concept of binary potential
in units of negative, squared speed of light in vacuum. Accordingly, the event
horizon is a fundamental holographic sphere in thermodynamic equilibrium with
only one exterior side: a noncompressible binary message that maximizes Shannon
entropy. Therefore, the Jordan-Brouwer separation theorem and generalized
Stokes theorem do not hold for black holes. We introduce the concept of
inertial potential and demonstrate its equivalence to the variational
potential, which ensures that any inertial acceleration represents a
nonequilibrium thermodynamic condition. We introduce the concept of the
complementary time period and relate it with the classical time period through
integral powers of the imaginary unit to formulate the notions of unobservable
velocity and acceleration, which are perpendicular and tangential to the
holographic sphere, respectively, and bound with the observable velocity and
acceleration based on Pythagorean relations. We further discuss certain
dynamics scenarios between the two masses. The concept of black hole
informationless emission is introduced as a complement to informationless
Bekenstein absorption and extended to arbitrary wavelengths. Black hole quantum
statistics with degeneracy interpreted as the number of Planck areas on the
event horizon are discussed. The study concludes that holographic screens and
equipotential surfaces are spherical equivalents, and every observer is a
sphere in nonequilibrium thermodynamic condition. Lastly, we propose a solution
to the black hole information paradox.
| arxiv topic:physics.gen-ph |
arxiv_dataset-119511910.11181 | The Measure Game
math.LO
We study a game first introduced by Martin (actually we use a slight
variation of this game) which plays a role for measure analogous to the
Banach-Mazur game for category. We first present proofs for the basic
connections between this game and measure, and then use the game to prove
fundamental measure theoretic results such as Fubini's theorem, the
Borel-Cantelli lemma, and a general unfolding result for the game which gives,
for example, the measurability of $\boldsymbol{\Sigma}^1_1$ sets. We also use
the game to give a new, more constructive, proof of a strong form of the
R\'{e}nyi-Lamperti lemma, an important result in probability theory with many
applications to number theory. The proofs we give are all direct combinatorial
arguments using the game, and do not depend on known measure theoretic
arguments.
| arxiv topic:math.LO |
arxiv_dataset-119521910.11281 | Ultrashort waveguide tapers based on Luneburg lens
physics.optics physics.app-ph
In integrated photonic circuits, silicon-on-insulator waveguides with
different geometries have been employed to realize a variety of components.
Therefore, efficient coupling of two different waveguides is crucial. In this
paper, focusing property of the Luneburg lens is exploited to design waveguide
tapers. The Luneburg lens, truncated in a shape of a parabolic taper with
reduced footprint, is utilized to connect a 10 $\mu$m-wide waveguide to a 0.5
$\mu$m one with the same thickness with an average coupling loss of 0.35 dB in
the entire O, E, S, C, L, and U bands of optical communications. The proposed
compact taper with the length of 11 $\mu$m is implemented by varying the
thickness of the guiding layer and compared with three conventional tapers with
the same length. However, designing a coupler to connect waveguides with
different thicknesses and widths is more challenging. By applying
quasi-conformal transformation optics, we flatten the Luneburg lens and
consequently increase the refractive index on the flattened side. As a result,
we are able to couple two waveguides with different thicknesses and widths. The
numerical simulations are used to evaluate the theoretically designed tapers.
To our knowledge, this is the first study presenting ultrashort tapers based on
truncated Luneburg lens.
| arxiv topic:physics.optics physics.app-ph |
arxiv_dataset-119531910.11381 | Superstatistics of Schr\"odinger Equation with Pseudoharmonic potential
in an External Magnetic and Aharanov-Bohm(AB) Fields
quant-ph
In this work, the thermodynamic property of pseudoharmonic potential in the
presence of external magnetic and AB fields is investigated. We used effective
Boltzmann factor within the superstatistics formalism to obtain the
thermodynamic properties such as Helmholtz free energy (F), Internal energy
(U), entropy(S) and specific heat (C) of the system. In addition, we discuss
the result of the thermodynamic properties of some selected diatomic molecules
of N2, Cl2, I2 and CH using their experimental spectroscopic parameters and
that of the variation of the deformation parameter of q=0,0.3,0.7. We also
illustrated with some graphs for clarity of our results in both cases.
| arxiv topic:quant-ph |
arxiv_dataset-119541910.11481 | Multimodal Image Outpainting With Regularized Normalized Diversification
cs.CV
In this paper, we study the problem of generating a set ofrealistic and
diverse backgrounds when given only a smallforeground region. We refer to this
task as image outpaint-ing. The technical challenge of this task is to
synthesize notonly plausible but also diverse image outputs.
Traditionalgenerative adversarial networks suffer from mode collapse.While
recent approaches propose to maximize orpreserve the pairwise distance between
generated sampleswith respect to their latent distance, they do not
explicitlyprevent the diverse samples of different conditional inputsfrom
collapsing. Therefore, we propose a new regulariza-tion method to encourage
diverse sampling in conditionalsynthesis. In addition, we propose a feature
pyramid dis-criminator to improve the image quality. Our experimen-tal results
show that our model can produce more diverseimages without sacrificing visual
quality compared to state-of-the-arts approaches in both the CelebA face
dataset and the Cityscape scene dataset.
| arxiv topic:cs.CV |
arxiv_dataset-119551910.11581 | Mesoscopic Stoner instability in open quantum dots: suppression of
Coleman-Weinberg mechanism by electron tunneling
cond-mat.mes-hall cond-mat.str-el
The mesoscopic Stoner instability is an intriguing manifestation of symmetry
breaking in isolated metallic quantum dots, underlined by the competition
between single-particle energy and Heisenberg exchange interaction. Here we
study this phenomenon in the presence of tunnel coupling to a reservoir. We
analyze the spin susceptibility of electrons on the quantum dot for different
values of couplings and temperature. Our results indicate the existence of a
quantum phase transition at a critical value of the tunneling coupling, which
is determined by the Stoner-enhanced exchange interaction. This quantum phase
transition is a manifestation of the suppression of the Coleman-Weinberg
mechanism of symmetry breaking, induced by coupling to the reservoir.
| arxiv topic:cond-mat.mes-hall cond-mat.str-el |
arxiv_dataset-119561910.11681 | Higher pullbacks of modular forms on orthogonal groups
math.NT
We apply differential operators to modular forms on orthogonal groups
$\mathrm{O}(2, \ell)$ to construct infinite families of modular forms on
special cycles. These operators generalize the quasi-pullback. The subspaces of
theta lifts are preserved; in particular, the higher pullbacks of the lift of a
(lattice-index) Jacobi form $\phi$ are theta lifts of partial development
coefficients of $\phi$. For certain lattices of signature (2, 2) and (2, 3),
for which there are interpretations as Hilbert-Siegel modular forms, we observe
that the higher pullbacks coincide with differential operators introduced by
Cohen and Ibukiyama.
| arxiv topic:math.NT |
arxiv_dataset-119571910.11781 | Automorphism orbits and element orders in finite groups:
almost-solubility and the Monster
math.GR
For a finite group $G$, we denote by $\omega(G)$ the number of
$\operatorname{Aut}(G)$-orbits on $G$, and by $\operatorname{o}(G)$ the number
of distinct element orders in $G$. In this paper, we are primarily concerned
with the two quantities $\mathfrak{d}(G):=\omega(G)-\operatorname{o}(G)$ and
$\mathfrak{q}(G):=\omega(G)/\operatorname{o}(G)$, each of which may be viewed
as a measure for how far $G$ is from being an AT-group in the sense of Zhang
(that is, a group with $\omega(G)=\operatorname{o}(G)$). We show that the index
$|G:\operatorname{Rad}(G)|$ of the soluble radical $\operatorname{Rad}(G)$ of
$G$ can be bounded from above both by a function in $\mathfrak{d}(G)$ and by a
function in $\mathfrak{q}(G)$ and $\operatorname{o}(\operatorname{Rad}(G))$. We
also obtain a curious quantitative characterisation of the Fischer-Griess
Monster group $\operatorname{M}$.
| arxiv topic:math.GR |
arxiv_dataset-119581910.11881 | Short-distance constraints on hadronic light-by-light scattering in the
anomalous magnetic moment of the muon
hep-ph hep-ex hep-lat nucl-th
A key ingredient in the evaluation of hadronic light-by-light (HLbL)
scattering in the anomalous magnetic moment of the muon $(g-2)_\mu$ concerns
short-distance constraints (SDCs) that follow from QCD by means of the operator
product expansion. Here we concentrate on the most important such constraint,
in the longitudinal amplitudes, and show that it can be implemented efficiently
in terms of a Regge sum over excited pseudoscalar states, constrained by
phenomenological input on masses, two-photon couplings, as well as SDCs on HLbL
scattering and the pseudoscalar transition form factors (TFFs). Our estimate of
the effect of the longitudinal SDCs on the HLbL contribution is: $\Delta
a_\mu^\text{LSDC}=13(6)\times 10^{-11}$. This is significantly smaller than
previous estimates, which mostly relied on an ad-hoc modification of the
pseudoscalar poles and led to up to a $40\%$ increase with respect to the
nominal pseudoscalar-pole contributions, when evaluated with modern input for
the relevant TFFs. We also comment on the status of the transversal SDCs and,
by matching to perturbative QCD, argue that the corresponding correction will
be significantly smaller than its longitudinal counterpart.
| arxiv topic:hep-ph hep-ex hep-lat nucl-th |
arxiv_dataset-119591910.11981 | Novel Co-variant Feature Point Matching Based on Gaussian Mixture Model
cs.CV
The feature frame is a key idea of feature matching problem between two
images. However, most of the traditional matching methods only simply employ
the spatial location information (the coordinates), which ignores the shape and
orientation information of the local feature. Such additional information can
be obtained along with coordinates using general co-variant detectors such as
DOG, Hessian, Harris-Affine and MSER. In this paper, we develop a novel method
considering all the feature center position coordinates, the local feature
shape and orientation information based on Gaussian Mixture Model for
co-variant feature matching. We proposed three sub-versions in our method for
solving the matching problem in different conditions: rigid, affine and
non-rigid, respectively, which all optimized by expectation maximization
algorithm. Due to the effective utilization of the additional shape and
orientation information, the proposed model can significantly improve the
performance in terms of convergence speed and recall. Besides, it is more
robust to the outliers.
| arxiv topic:cs.CV |
arxiv_dataset-119601910.12081 | A computationally efficient robust model predictive control framework
for uncertain nonlinear systems -- extended version
eess.SY cs.SY
In this paper, we present a nonlinear robust model predictive control (MPC)
framework for general (state and input dependent) disturbances. This approach
uses an online constructed tube in order to tighten the nominal (state and
input) constraints. To facilitate an efficient online implementation, the shape
of the tube is based on an offline computed incremental Lyapunov function with
a corresponding (nonlinear) incrementally stabilizing feedback. Crucially, the
online optimization only implicitly includes these nonlinear functions in terms
of scalar bounds, which enables an efficient implementation. Furthermore, to
account for an efficient evaluation of the worst case disturbance, a simple
function is constructed offline that upper bounds the possible disturbance
realizations in a neighbourhood of a given point of the open-loop trajectory.
The resulting MPC scheme ensures robust constraint satisfaction and practical
asymptotic stability with a moderate increase in the online computational
demand compared to a nominal MPC. We demonstrate the applicability of the
proposed framework in comparison to state of the art robust MPC approaches with
a nonlinear benchmark example. This paper is an extended version of [1], and
contains further details and additional considers: continuous-time systems
(App. A), more general nonlinear constraints (App. B) and special cases (Sec.
IV).
| arxiv topic:eess.SY cs.SY |
arxiv_dataset-119611910.12181 | Multi-source Domain Adaptation for Semantic Segmentation
cs.CV cs.LG eess.IV
Simulation-to-real domain adaptation for semantic segmentation has been
actively studied for various applications such as autonomous driving. Existing
methods mainly focus on a single-source setting, which cannot easily handle a
more practical scenario of multiple sources with different distributions. In
this paper, we propose to investigate multi-source domain adaptation for
semantic segmentation. Specifically, we design a novel framework, termed
Multi-source Adversarial Domain Aggregation Network (MADAN), which can be
trained in an end-to-end manner. First, we generate an adapted domain for each
source with dynamic semantic consistency while aligning at the pixel-level
cycle-consistently towards the target. Second, we propose sub-domain
aggregation discriminator and cross-domain cycle discriminator to make
different adapted domains more closely aggregated. Finally, feature-level
alignment is performed between the aggregated domain and target domain while
training the segmentation network. Extensive experiments from synthetic GTA and
SYNTHIA to real Cityscapes and BDDS datasets demonstrate that the proposed
MADAN model outperforms state-of-the-art approaches. Our source code is
released at: https://github.com/Luodian/MADAN.
| arxiv topic:cs.CV cs.LG eess.IV |
arxiv_dataset-119621910.12281 | Deep convolutional autoencoder for cryptocurrency market analysis
cs.LG q-fin.ST stat.ML
This study attempts to analyze patterns in cryptocurrency markets using a
special type of deep neural networks, namely a convolutional autoencoder. The
method extracts the dominant features of market behavior and classifies the 40
studied cryptocurrencies into several classes for twelve 6-month periods
starting from 15th May 2013. Transitions from one class to another with time
are related to the maturement of cryptocurrencies. In speculative
cryptocurrency markets, these findings have potential implications for
investment and trading strategies.
| arxiv topic:cs.LG q-fin.ST stat.ML |
arxiv_dataset-119631910.12381 | Transferring neural speech waveform synthesizers to musical instrument
sounds generation
eess.AS cs.SD stat.ML
Recent neural waveform synthesizers such as WaveNet, WaveGlow, and the
neural-source-filter (NSF) model have shown good performance in speech
synthesis despite their different methods of waveform generation. The
similarity between speech and music audio synthesis techniques suggests
interesting avenues to explore in terms of the best way to apply speech
synthesizers in the music domain. This work compares three neural synthesizers
used for musical instrument sounds generation under three scenarios: training
from scratch on music data, zero-shot learning from the speech domain, and
fine-tuning-based adaptation from the speech to the music domain. The results
of a large-scale perceptual test demonstrated that the performance of three
synthesizers improved when they were pre-trained on speech data and fine-tuned
on music data, which indicates the usefulness of knowledge from speech data for
music audio generation. Among the synthesizers, WaveGlow showed the best
potential in zero-shot learning while NSF performed best in the other scenarios
and could generate samples that were perceptually close to natural audio.
| arxiv topic:eess.AS cs.SD stat.ML |
arxiv_dataset-119641910.12481 | Generative Well-intentioned Networks
cs.LG stat.ML
We propose Generative Well-intentioned Networks (GWINs), a novel framework
for increasing the accuracy of certainty-based, closed-world classifiers. A
conditional generative network recovers the distribution of observations that
the classifier labels correctly with high certainty. We introduce a reject
option to the classifier during inference, allowing the classifier to reject an
observation instance rather than predict an uncertain label. These rejected
observations are translated by the generative network to high-certainty
representations, which are then relabeled by the classifier. This architecture
allows for any certainty-based classifier or rejection function and is not
limited to multilayer perceptrons. The capability of this framework is assessed
using benchmark classification datasets and shows that GWINs significantly
improve the accuracy of uncertain observations.
| arxiv topic:cs.LG stat.ML |
arxiv_dataset-119651910.12581 | A Multivariate Elo-based Learner Model for Adaptive Educational Systems
cs.CY cs.HC
The Elo rating system has been recognised as an effective method for
modelling students and items within adaptive educational systems. The existing
Elo-based models have the limiting assumption that items are only tagged with a
single concept and are mainly studied in the context of adaptive testing
systems. In this paper, we introduce a multivariate Elo-based learner model
that is suitable for the domains where learning items can be tagged with
multiple concepts, and investigate its fit in the context of adaptive learning.
To evaluate the model, we first compare the predictive performance of the
proposed model against the standard Elo-based model using synthetic and public
data sets. Our results from this study indicate that our proposed model has
superior predictive performance compared to the standard Elo-based model, but
the difference is rather small. We then investigate the fit of the proposed
multivariate Elo-based model by integrating it into an adaptive learning system
which incorporates the principles of open learner models (OLMs). The results
from this study suggest that the availability of additional parameters derived
from multivariate Elo-based models have two further advantages: guiding
adaptive behaviour for the system and providing additional insight for students
and instructors.
| arxiv topic:cs.CY cs.HC |
arxiv_dataset-119661910.12681 | Local well-posedness for the quadratic Schrodinger equation in
two-dimensional compact manifolds with boundary
math.AP
We consider the quadractic NLS posed on a bidimensional compact Riemannian
manifold $(M, g)$ with $ \partial M \neq \emptyset$. Using bilinear and
gradient bilinear Strichartz estimates for Schr\"odinger operators in
two-dimensional compact manifolds proved by J. Jiang in \cite{JIANG} we deduce
a new evolution bilinear estimates. Consequently, using Bourgain's spaces, we
obtain a local well-posedness result for given data $u_0\in H^s(M)$ whenever
$s> \frac{2}{3}$ in such manifolds.
| arxiv topic:math.AP |
arxiv_dataset-119671910.12781 | Empirical Analysis of Session-Based Recommendation Algorithms
cs.IR cs.LG cs.NE
Recommender systems are tools that support online users by pointing them to
potential items of interest in situations of information overload. In recent
years, the class of session-based recommendation algorithms received more
attention in the research literature. These algorithms base their
recommendations solely on the observed interactions with the user in an ongoing
session and do not require the existence of long-term preference profiles. Most
recently, a number of deep learning based ("neural") approaches to
session-based recommendations were proposed. However, previous research
indicates that today's complex neural recommendation methods are not always
better than comparably simple algorithms in terms of prediction accuracy.
With this work, our goal is to shed light on the state-of-the-art in the area
of session-based recommendation and on the progress that is made with neural
approaches. For this purpose, we compare twelve algorithmic approaches, among
them six recent neural methods, under identical conditions on various datasets.
We find that the progress in terms of prediction accuracy that is achieved with
neural methods is still limited. In most cases, our experiments show that
simple heuristic methods based on nearest-neighbors schemes are preferable over
conceptually and computationally more complex methods. Observations from a user
study furthermore indicate that recommendations based on heuristic methods were
also well accepted by the study participants. To support future progress and
reproducibility in this area, we publicly share the session-rec evaluation
framework that was used in our research.
| arxiv topic:cs.IR cs.LG cs.NE |
arxiv_dataset-119681910.12881 | The Null Geodesics of the Kerr Exterior
gr-qc astro-ph.HE hep-th
The null geodesic equation in the Kerr spacetime can be expressed as a set of
integral equations involving certain potentials. We classify the roots of these
potentials and express the integrals in manifestly real Legendre elliptic form.
We then solve the equations using Jacobi elliptic functions, providing the
complete set of null geodesics of the Kerr exterior as explicit parameterized
curves.
| arxiv topic:gr-qc astro-ph.HE hep-th |
arxiv_dataset-119691910.12981 | Non-integrability on AdS$_3$ supergravity backgrounds
hep-th
We investigate classical integrability on two recently discovered classes of
backgrounds in massive IIA supergravity. These vacua are of the form
AdS$_3\times\,$S$^2\times\mathbb{R}\times\,$CY$_2$, they preserve small
$\mathcal{N}=(0,4)$ supersymmetry and are associated with D8$-$D6$-$D4$-$D2
Hanany-Witten brane set-ups. We choose an appropriate string embedding and use
differential Galois theory on its associated Hamiltonian system, intending to
produce the conditions under which Liouvillian solutions may occur. By
constraining the parameters of the system according to the consistency of the
associate brane set-ups we prove that no such conditions exist, yielding the
complete non-integrability of these vacua. That is, up to the trivial cases
where the background reduces to the Abelian and non-Abelian T-dual of
AdS$_3\times\,$S$^3\times\,$T$^4$.
| arxiv topic:hep-th |
arxiv_dataset-119701910.13081 | Classification Calibration for Long-tail Instance Segmentation
cs.CV
Remarkable progress has been made in object instance detection and
segmentation in recent years. However, existing state-of-the-art methods are
mostly evaluated with fairly balanced and class-limited benchmarks, such as
Microsoft COCO dataset [8]. In this report, we investigate the performance drop
phenomenon of state-of-the-art two-stage instance segmentation models when
processing extreme long-tail training data based on the LVIS [5] dataset, and
find a major cause is the inaccurate classification of object proposals. Based
on this observation, we propose to calibrate the prediction of classification
head to improve recognition performance for the tail classes. Without much
additional cost and modification of the detection model architecture, our
calibration method improves the performance of the baseline by a large margin
on the tail classes. Codes will be available. Importantly, after the
submission, we find significant improvement can be further achieved by
modifying the calibration head, which we will update later.
| arxiv topic:cs.CV |
arxiv_dataset-119711910.13181 | Bridging the ELBO and MMD
cs.LG stat.ML
One of the challenges in training generative models such as the variational
auto encoder (VAE) is avoiding posterior collapse. When the generator has too
much capacity, it is prone to ignoring latent code. This problem is exacerbated
when the dataset is small, and the latent dimension is high. The root of the
problem is the ELBO objective, specifically the Kullback-Leibler (KL)
divergence term in objective function \citep{zhao2019infovae}. This paper
proposes a new objective function to replace the KL term with one that emulates
the maximum mean discrepancy (MMD) objective. It also introduces a new
technique, named latent clipping, that is used to control distance between
samples in latent space. A probabilistic autoencoder model, named $\mu$-VAE, is
designed and trained on MNIST and MNIST Fashion datasets, using the new
objective function and is shown to outperform models trained with ELBO and
$\beta$-VAE objective. The $\mu$-VAE is less prone to posterior collapse, and
can generate reconstructions and new samples in good quality. Latent
representations learned by $\mu$-VAE are shown to be good and can be used for
downstream tasks such as classification.
| arxiv topic:cs.LG stat.ML |
arxiv_dataset-119721910.13281 | Geometric Flows of Curves, Two-Component Camassa-Holm Equation and
Generalized Heisenberg Ferromagnet Equation
nlin.SI
In this paper, we study the generalized Heisenberg ferromagnet equation,
namely, the M-CVI equation. This equation is integrable. The integrable motion
of the space curves induced by the M-CVI equation is presented. Using this
result, the Lakshmanan (geometrical) equivalence between the M-CVI equation and
the two-component Camassa-Holm equation is established. Note that these
equations are gauge equivalent each to other.
| arxiv topic:nlin.SI |
arxiv_dataset-119731910.13381 | Local Wiener's Theorem and Coherent Sets of Frequencies
math.CA
Using a local analog of the Wiener-Levi theorem, we investigate the class of
measures on Euclidean space with discrete support and spectrum. Also, we find a
new sufficient conditions for a discrete set in Euclidean space to be a
coherent set of frequencies.
| arxiv topic:math.CA |
arxiv_dataset-119741910.13481 | Classical nucleation theory predicts the shape of the nucleus in
homogeneous solidification
cond-mat.stat-mech
Macroscopic models of nucleation provide powerful tools for understanding
activated phase transition processes. These models do not provide atomistic
insights and can thus sometime lack material-specific descriptions. Here we
provide a comprehensive framework for constructing a continuum picture from an
atomistic simulation of homogeneous nucleation. We use this framework to
determine the shape of the equilibrium solid nucleus that forms inside bulk
liquid for a Lennard-Jones potential. From this shape, we then extract the
anisotropy of the solid-liquid interfacial free energy, by performing a reverse
Wulff construction in the space of spherical harmonic expansions. We find that
the shape of the nucleus is nearly spherical and that its anisotropy can be
perfectly described using classical models.
| arxiv topic:cond-mat.stat-mech |
arxiv_dataset-119751910.13581 | Fission in a microscopic framework: from basic science to support for
applications
nucl-th nucl-ex
Recent developments, both in theoretical modeling and computational power,
have allowed us to make progress on a goal not fully achieved yet in nuclear
theory: a microscopic theory of nuclear fission. Even if the complete
microscopic description remains a computationally demanding task, the
information that can be provided by current calculations can be extremely
useful to guide and constrain more phenomenological approaches, which are
simpler to implement. First, a microscopic model that describes the real-time
dynamics of the fissioning system can justify or rule out some of the
approximations. Second, the microscopic approach can be used to obtain trends,
e.g., with increasing excitation energy of the fissioning system, or even to
compute observables that cannot be otherwise calculated in phenomenological
approaches or that can be hindered by the limitations of the method. We briefly
present in this contribution the time-dependent superfluid local density
approximation (TDSLDA) approach to nuclear fission, approach that has become a
very successful theoretical model in many areas of many-body research. The
TDSLDA incorporates the effects of the continuum, the dynamics of the pairing
field, and the numerical solution is implemented with controlled approximations
and negligible numerical corrections. The main part of the current contribution
will be dedicated to discussing the method, and recent results concerning the
fission dynamics. In addition, we present results on the excitation energy
sharing between the fragments, which are in agreement with a qualitative
conclusions extracted from a limited number of experimental measurements of
properties of prompt neutrons.
| arxiv topic:nucl-th nucl-ex |
arxiv_dataset-119761910.13681 | The Domain Shift Problem of Medical Image Segmentation and
Vendor-Adaptation by Unet-GAN
eess.IV cs.CV
Convolutional neural network (CNN), in particular the Unet, is a powerful
method for medical image segmentation. To date Unet has demonstrated
state-of-art performance in many complex medical image segmentation tasks,
especially under the condition when the training and testing data share the
same distribution (i.e. come from the same source domain). However, in clinical
practice, medical images are acquired from different vendors and centers. The
performance of a U-Net trained from a particular source domain, when
transferred to a different target domain (e.g. different vendor, acquisition
parameter), can drop unexpectedly. Collecting a large amount of annotation from
each new domain to retrain the U-Net is expensive, tedious, and practically
impossible. In this work, we proposed a generic framework to address this
problem, consisting of (1) an unpaired generative adversarial network (GAN) for
vendor-adaptation, and (2) a Unet for object segmentation. In the proposed
Unet-GAN architecture, GAN learns from Unet at the feature level that is
segmentation-specific. We used cardiac cine MRI as the example, with three
major vendors (Philips, Siemens, and GE) as three domains, while the
methodology can be extended to medical images segmentation in general. The
proposed method showed significant improvement of the segmentation results
across vendors. The proposed Unet-GAN provides an annotation-free solution to
the cross-vendor medical image segmentation problem, potentially extending a
trained deep learning model to multi-center and multi-vendor use in real
clinical scenario.
| arxiv topic:eess.IV cs.CV |
arxiv_dataset-119771910.13781 | Classification of irreducible modules for Bershadsky-Polyakov algebra at
certain levels
math.QA math-ph math.MP math.RT
We study the representation theory of the Bershadsky-Polyakov algebra
$\mathcal W_k = \mathcal{W}_k(sl_3,f_{\theta})$. In particular, Zhu algebra of
$\mathcal W_k$ is isomorphic to a certain quotient of the Smith algebra, after
changing the Virasoro vector. We classify all modules in the category
$\mathcal{O}$ for the Bershadsky-Polyakov algebra $\mathcal W_k$ when $k=-5/3,
-9/4, -1,0$. In the case $k=0$ we show that the Zhu algebra $A(\mathcal W_k)$
has $2$--dimensional indecomposable modules.
| arxiv topic:math.QA math-ph math.MP math.RT |
arxiv_dataset-119781910.13881 | Normal limit laws for vertex degrees in randomly grown hooking networks
and bipolar networks
math.PR math.CO
We consider two types of random networks grown in blocks. Hooking networks
are grown from a set of graphs as blocks, each with a labelled vertex called a
hook. At each step in the growth of the network, a vertex called a latch is
chosen from the hooking network and a copy of one of the blocks is attached by
fusing its hook with the latch. Bipolar networks are grown from a set of
directed graphs as blocks, each with a single source and a single sink. At each
step in the growth of the network, an arc is chosen and is replaced with a copy
of one of the blocks. Using P\'olya urns, we prove normal limit laws for the
degree distributions of both networks. We extend previous results by allowing
for more than one block in the growth of the networks and by studying
arbitrarily large degrees.
| arxiv topic:math.PR math.CO |
arxiv_dataset-119791910.13981 | The 3D structure of CO depletion in high-mass prestellar regions
astro-ph.GA
Disentangling the different stages of the star-formation process, in
particular in the high-mass regime, is a challenge in astrophysics. Chemical
clocks could help alleviating this problem, but their evolution strongly
depends on many parameters, leading to degeneracy in the interpretation of the
observational data. One of these uncertainties is the degree of CO depletion.
We present here the first self-consistent magneto-hydrodynamic simulations of
high-mass star-forming regions at different scales, fully coupled with a
non-equilibrium chemical network, which includes C-N-O bearing molecules.
Depletion and desorption processes are treated time-dependently. The results
show that full CO-depletion (i.e. all gas-phase CO frozen-out on the surface of
dust grains), can be reached very quickly, in one third or even smaller
fractions of the free-fall time, whether the collapse proceeds on slow or fast
timescales. This leads to a high level of deuteration in a short time both for
typical tracers like N$_2$H$^+$, as well as for the main ion H$_3^+$, the
latter being in general larger and more extended. N$_2$ depletion is slightly
less efficient, and no direct effects on N-bearing molecules and deuterium
fractionation are observed. We show that CO depletion is not the only driver of
deuteration, and that there is a strong impact on $D_{frac}$ when changing the
grain-size. We finally apply a two-dimensional gaussian Point Spread Function
to our results to mimic observations with single-dish and interferometers. Our
findings suggest that the low-values observed in high-mass star-forming clumps
are in reality masking a full-depletion stage in the inner 0.1 pc region.
| arxiv topic:astro-ph.GA |
arxiv_dataset-119801910.14081 | Duality and Stability in Complex Multiagent State-Dependent Network
Dynamics
eess.SY cs.MA cs.SY math.DS math.OC
Despite significant progress on stability analysis of conventional multiagent
networked systems with weakly coupled state-network dynamics, most of the
existing results have shortcomings in addressing multiagent systems with highly
coupled state-network dynamics. Motivated by numerous applications of such
dynamics, in our previous work [1], we initiated a new direction for stability
analysis of such systems that uses a sequential optimization framework.
Building upon that, in this paper, we extend our results by providing another
angle on multiagent network dynamics from a duality perspective, which allows
us to view the network structure as dual variables of a constrained nonlinear
program. Leveraging that idea, we show that the evolution of the coupled
state-network multiagent dynamics can be viewed as iterates of a primal-dual
algorithm for a static constrained optimization/saddle-point problem. This view
bridges the Lyapunov stability of state-dependent network dynamics and
frequently used optimization techniques such as block coordinated descent,
mirror descent, the Newton method, and the subgradient method. As a result, we
develop a systematic framework for analyzing the Lyapunov stability of
state-dependent network dynamics using techniques from nonlinear optimization.
Finally, we support our theoretical results through numerical simulations from
social science.
| arxiv topic:eess.SY cs.MA cs.SY math.DS math.OC |
arxiv_dataset-119811910.14181 | Irregularity of distribution in Wasserstein distance
math.CA
We study the non-uniformity of probability measures on the interval and the
circle. On the interval, we identify the Wasserstein-$p$ distance with the
classical $L^p$-discrepancy. We thereby derive sharp estimates in Wasserstein
distances for the irregularity of distribution of sequences on the interval and
the circle. Furthermore, we prove an $L^p$-adapted Erd\H{o}s$-$Tur\'{a}n
inequality.
| arxiv topic:math.CA |
arxiv_dataset-119821910.14281 | Probing QCD critical fluctuations from the yield ratio of strange
hadrons in relativistic heavy-ion collisions
hep-ph nucl-ex nucl-th
By analyzing the available data on strange hadrons in central Pb+Pb
collisions from the NA49 Collaboration at the Super Proton Synchrotron (SPS)
and in central Au+Au collisions from the STAR Collaboration at the Relativistic
Heavy-Ion Collider (RHIC) in a wide collision energy range from $\sqrt{s_{\rm
NN}}$ = 6.3 GeV to 200 GeV, we find a possible non-monotonic behavior in the
ratio $\mathcal{O}_\text{K-$\Xi$-$\phi$-$\Lambda$}$=
$\frac{N(K^+)N(\Xi^-)}{N(\phi)N(\Lambda)}$ of $K^+$, $\Xi^-$, $\phi$, and
$\Lambda$ yields as a function of $\sqrt{s_{\rm NN}}$. Based on the quark
coalescence model, which can take into account the effect of quark density
fluctuations on hadron production, a possible non-monotonic behavior in the
dependence of the strange quark density fluctuation on $\sqrt{s_{NN}}$ is
obtained. This is in contrast to the coalescence model that does not include
quark density fluctuations and also to the statistical hadronization model as
both fail to describe even qualitatively the collision energy dependence of the
ratio $\mathcal{O}_\text{K-$\Xi$-$\phi$-$\Lambda$}$. Our findings thus suggest
that the signal and location of a possible critical endpoint in the QCD phase
diagram, which is expected to result in large quark density fluctuations, can
be found in the on-going Bean Energy Scan program at RHIC.
| arxiv topic:hep-ph nucl-ex nucl-th |
arxiv_dataset-119831910.14381 | A note on commutative Kleene algebra
cs.FL cs.LO
In this paper we present a detailed proof of an important result of algebraic
logic: namely that the free commutative Kleene algebra is the space of
semilinear sets. The first proof of this result was proposed by Redko in 1964,
and simplified and corrected by Pilling in his 1970 thesis. However, we feel
that a new account of this proof is needed now. This result has acquired a
particular importance in recent years, since it is a key component in the
completeness proofs of several algebraic models of concurrent computations
(bi-Kleene algebra, concurrent Kleene algebra...). To that effect, we present a
new proof of this result.
| arxiv topic:cs.FL cs.LO |
arxiv_dataset-119841910.14481 | Continual Unsupervised Representation Learning
cs.LG cs.AI cs.CV stat.ML
Continual learning aims to improve the ability of modern learning systems to
deal with non-stationary distributions, typically by attempting to learn a
series of tasks sequentially. Prior art in the field has largely considered
supervised or reinforcement learning tasks, and often assumes full knowledge of
task labels and boundaries. In this work, we propose an approach (CURL) to
tackle a more general problem that we will refer to as unsupervised continual
learning. The focus is on learning representations without any knowledge about
task identity, and we explore scenarios when there are abrupt changes between
tasks, smooth transitions from one task to another, or even when the data is
shuffled. The proposed approach performs task inference directly within the
model, is able to dynamically expand to capture new concepts over its lifetime,
and incorporates additional rehearsal-based techniques to deal with
catastrophic forgetting. We demonstrate the efficacy of CURL in an unsupervised
learning setting with MNIST and Omniglot, where the lack of labels ensures no
information is leaked about the task. Further, we demonstrate strong
performance compared to prior art in an i.i.d setting, or when adapting the
technique to supervised tasks such as incremental class learning.
| arxiv topic:cs.LG cs.AI cs.CV stat.ML |
arxiv_dataset-119851910.14581 | Stark-Heegner cycles attached to Bianchi modular forms
math.NT
Let f be a Bianchi modular form, that is, an automorphic form for GL(2) over
an imaginary quadratic field F, and let P be a prime of F at which f is new.
Let K be a quadratic extension of F, and L(f/K,s) the L-function of the
base-change of f to K. Under certain hypotheses on f and K, the functional
equation of L(f/K,s) ensures that it vanishes at the central point. The
Bloch--Kato conjecture predicts that this should force the existence of
non-trivial classes in an appropriate global Selmer group attached to f and K.
In this paper, we use the theory of double integrals developed by Barrera
Salazar and the second author to construct certain P-adic Abel--Jacobi maps,
which we use to propose a construction of such classes via "Stark--Heegner
cycles". This builds on ideas of Darmon and in particular generalises an
approach of Rotger and Seveso in the setting of classical modular forms.
| arxiv topic:math.NT |
arxiv_dataset-119861911.00006 | From veering triangulations to link spaces and back again
math.GT
This paper is the third in a sequence establishing a dictionary between the
combinatorics of veering triangulations equipped with appropriate filling
slopes, and the dynamics of pseudo-Anosov flows (without perfect fits) on
closed three-manifolds.
Our motivation comes from the work of Agol and Gu\'eritaud. Agol introduced
veering triangulations of mapping tori as a tool for understanding the surgery
parents of pseudo-Anosov mapping tori. Gu\'eritaud gave a new construction of
veering triangulations of mapping tori using the orbit spaces of their
suspension flows. Generalising this, Agol and Gu\'eritaud announced a method
that, given a closed manifold with a pseudo-Anosov flow (without perfect fits),
produces a veering triangulation equipped with filling slopes.
In this paper we build, from a veering triangulation, a canonical circular
order on the cusps of the universal cover. Using this we build the veering
circle and the link space. These are the first entries in the promised
dictionary. The link space and the circle are, respectively, analogous to the
orbit space of a flow and to Fenley's boundary at infinity of the orbit space.
In the other direction, and using our previous work, we prove that the
veering triangulation is recovered (up to canonical isomorphism) from the
dynamics of the fundamental group acting on the link space. This is the first
step in proving that our dictionary gives a bijection between the two theories.
| arxiv topic:math.GT |
arxiv_dataset-119871911.00106 | An efficient algorithm of solution for the flow of generalized Newtonian
fluid in channels of simple geometries
physics.flu-dyn
In this paper a problem of stationary flow of generalized Newtonian fluid in
a thin channel is considered. An efficient algorithm of solution is proposed
that includes a flexible procedure for a continuous approximation of the
apparent viscosity by means of elementary functions combined with analytical
integration of the governing equations. The algorithm can be easily adapted to
circular or elliptic conduits. The accuracy and efficiency of computations are
analyzed using an example of the Carreau fluid. The proposed computational
scheme proves to be highly efficient and versatile providing excellent accuracy
of solution at a very low computational cost.
| arxiv topic:physics.flu-dyn |
arxiv_dataset-119881911.00206 | Symbolic extensions for 3-dimensional diffeomorphisms
math.DS
We prove that every $\mathcal{C}^{r}$ diffeomorphism with $r>1$ on a
three-dimensional manifold admits symbolic extensions, i.e. topological
extensions which are subshifts over a finite alphabet. This answers positively
a conjecture of Downarowicz and Newhouse in dimension three.
| arxiv topic:math.DS |
arxiv_dataset-119891911.00306 | Hawking's information puzzle: a solution realized in loop quantum
cosmology
gr-qc hep-th quant-ph
In approaches to quantum gravity, where smooth spacetime is an emergent
approximation of a discrete Planckian fundamental structure, any standard
effective field theoretical description will miss part of the degrees of
freedom and thus break unitarity. Here we show that these expectations can be
made precise in loop quantum cosmology. Concretely, even when loop quantum
cosmology is unitary at the fundamental level, when microscopic degrees of
freedom, irrelevant to low-energy cosmological observers, are suitably ignored,
pure states in the effective description evolve into mixed states due to
decoherence with the Planckian microscopic structure. When extrapolated to
black hole formation and evaporation, this concrete example provides a key
physical insight for a natural resolution of Hawking's information paradox.
| arxiv topic:gr-qc hep-th quant-ph |
arxiv_dataset-119901911.00406 | Formalizing the Dependency Pair Criterion for Innermost Termination
cs.LO cs.PL
Rewriting is a framework for reasoning about functional programming. The
dependency pair criterion is a well-known mechanism to analyze termination of
term rewriting systems. Functional specifications with an operational semantics
based on evaluation are related, in the rewriting framework, to the innermost
reduction relation. This paper presents a PVS formalization of the dependency
pair criterion for the innermost reduction relation: a term rewriting system is
innermost terminating if and only if it is terminating by the dependency pair
criterion. The paper also discusses the application of this criterion to check
termination of functional specifications.
| arxiv topic:cs.LO cs.PL |
arxiv_dataset-119911911.00506 | Six constructions of asymptotically optimal codebooks via the character
sums
cs.IT math.IT
In this paper, using additive characters of finite field, we find a codebook
which is equivalent to the measurement matrix in [20]. The advantage of our
construction is that it can be generalized naturally to construct the other
five classes of codebooks using additive and multiplicative characters of
finite field. We determine the maximal cross-correlation amplitude of these
codebooks by the properties of characters and character sums. We prove that all
the codebooks we constructed are asymptotically optimal with respect to the
Welch bound. The parameters of these codebooks are new.
| arxiv topic:cs.IT math.IT |
arxiv_dataset-119921911.00606 | The periodic integral orbits of polynomial recursions with integer
coefficients
math.DS
We show that polynomial recursions $x_{n+1}=x_{n}^{m}-k$ where $k,m$ are
integers and $m$ is positive have no nontrivial periodic integral orbits for
$m\geq3$. If $m=2$ then the recursion has integral two-cycles for infinitely
many values of $k$ but no higher period orbits. We also show that these
statements are true for all quadratic recursions.
| arxiv topic:math.DS |
arxiv_dataset-119931911.00706 | Layer breathing and shear modes in multilayer graphene: A DFT-vdW study
cond-mat.mes-hall cond-mat.mtrl-sci
In this work, we study structural and vibrational properties of multilayer
graphene using density-functional theory (DFT) with van der Waals (vdW)
functionals. Initially, we analyze how different vdW functionals compare by
evaluating the lattice parameters, elastic constants and vibrational
frequencies of low energy optical modes of graphite. Our results indicate that
the vdW-DF1-optB88 functional has the best overall performance on the
description of vibrational properties. Next, we use this functional to study
the influence of the vdW interactions on the structural and vibrational
properties of multilayer graphene. Specifically, we evaluate binding energies,
interlayer distances and phonon frequencies of layer breathing and shear modes.
We observe excellent agreement between our calculated results and available
experimental data, which suggests that this functional has truly predictive
power for layer-breathing and shear frequencies that have not been measured
yet. This indicates that careful selected vdW functionals can describe
interlayer bonding in graphene-related systems with good accuracy.
| arxiv topic:cond-mat.mes-hall cond-mat.mtrl-sci |
arxiv_dataset-119941911.00806 | Observation of quadratic Weyl points and double-helicoid arcs
cond-mat.mes-hall
Very recently, novel quasiparticles beyond those mimicking the elementary
high-energy particles such as Dirac and Weyl fermions have attracted great
interest in condensed matter physics and materials science1-9. Here we report
the first experimental observation of the long-desired quadratic Weyl points10
by using a three-dimensional chiral metacrystal of sound waves. Markedly
different from the newly observed unconventional quasiparticles5-9, such as the
spin-1 Weyl points and the charge-2 Dirac points that are featured respectively
with threefold and fourfold band crossings, the charge-2 Weyl points identified
here are simply twofold degenerate, and the dispersions around them are
quadratic in two directions and linear in the third one10. Besides the
essential nonlinear bulk dispersions, we further unveil the exotic
double-helicoid surface arcs that emanate from a projected quadratic Weyl point
and terminate at two projected conventional Weyl points through Fourier
transformation of the scanned surface fields. This unique global surface
connectivity provides conclusive evidence for the double topological charges of
such unconventional topological nodes.
| arxiv topic:cond-mat.mes-hall |
arxiv_dataset-119951911.00906 | Influences of weak disorder on dynamical quantum phase transitions of
anisotropic XY chain
cond-mat.dis-nn cond-mat.stat-mech
In this paper, the effects of disorder on the dynamical quantum phase
transitions (DQPTs) in the transverse-field anisotropic XY chain are studied by
numerically calculating the Loschmidt echo after quench. We obtain the formula
for calculating the Loschmidt echo of the inhomogeneous system in real space.
By comparing the results with that of the homogeneous chain, we find that when
the quench crosses the Ising transition, the small disorder will cause a new
critical point. As the disorder increases, more critical points of the DQPTs
will occur, constituting a critical region. In the quench across the
anisotropic transition, the disorder will cause a critical region near the
critical point, and the width of the critical region increases by the
disordered strength. In the case of quench passing through two critical lines,
the small disorder leads to the system to have three additional critical
points. When the quench is in the ferromagnetic phase, the large disorder
causes the two critical points of the homogeneous case to become a critical
region. And for the quench in the paramagnetic phase, the DQPTs will disappear
for large disorder.
| arxiv topic:cond-mat.dis-nn cond-mat.stat-mech |
arxiv_dataset-119961911.01006 | Fast Optical System Identification by Numerical Interferometry
eess.SP eess.IV
We propose a numerical interferometry method for identification of optical
multiply-scattering systems when only intensity can be measured. Our method
simplifies the calibration of optical transmission matrices from a quadratic to
a linear inverse problem by first recovering the phase of the measurements. We
show that by carefully designing the probing signals, measurement phase
retrieval amounts to a distance geometry problem---a multilateration---in the
complex plane. Since multilateration can be formulated as a small linear system
which is the same for entire rows of the transmission matrix, the phases can be
retrieved very efficiently. To speed up the subsequent estimation of
transmission matrices, we design calibration signals so as to take advantage of
the fast Fourier transform, achieving a numerical complexity almost linear in
the number of transmission matrix entries. We run experiments on real optical
hardware and use the numerically computed transmission matrix to recover an
unseen image behind a scattering medium. Where the previous state-of-the-art
method reports hours to compute the transmission matrix on a GPU, our method
takes only a few minutes on a CPU.
| arxiv topic:eess.SP eess.IV |
arxiv_dataset-119971911.01106 | Singular points detection with semantic segmentation networks
cs.CV
Singular points detection is one of the most classical and important problem
in the field of fingerprint recognition. However, current detection rates of
singular points are still unsatisfactory, especially for low-quality
fingerprints. Compared with traditional image processing-based detection
methods, methods based on deep learning only need the original fingerprint
image but not the fingerprint orientation field. In this paper, different from
other detection methods based on deep learning, we treat singular points
detection as a semantic segmentation problem and just use few data for
training. Furthermore, we propose a new convolutional neural network called
SinNet to extract the singular regions of interest and then use a blob
detection method called SimpleBlobDetector to locate the singular points. The
experiments are carried out on the test dataset from SPD2010, and the proposed
method has much better performance than the other advanced methods in most
aspects. Compared with the state-of-art algorithms in SPD2010, our method
achieves an increase of 11% in the percentage of correctly detected
fingerprints and an increase of more than 18% in the core detection rate.
| arxiv topic:cs.CV |
arxiv_dataset-119981911.01206 | Cardinal functions of purely atomic measures
math.CA
Let $\mu$ be a purely atomic measure. By
$f_\mu:[0,\infty)\to\{0,1,2,\dots,\omega,\mathfrak{c}\}$ we denote its cardinal
function $f_{\mu}(t)=\vert\{A\subset\mathbb N:\mu(A)=t\}\vert$. We study the
problem for which sets $R\subset\{0,1,2,\dots,\omega,\mathfrak{c}\}$ there is a
measure $\mu$ such that $R$ is the range of $f_\mu$. We are also interested in
the set-theoretic and topological properties of the set of $\mu$-values which
are obtained uniquely.
| arxiv topic:math.CA |
arxiv_dataset-119991911.01306 | Dynamic programming systems for modeling and control of the traffic in
transportation networks
math.OC math.DS
This thesis is entitled Dynamic programming systems for modeling and control
of the traffic in transportation networks. Two parts are distinguished in this
dissertation: 1) methods and approaches based on min-plus or max-plus algebra,
where the dynamics are deterministic dynamic programming systems; 2) methods
and approaches whose dynamic systems are non-linear but are interpreted as
stochastic dynamic programming systems. Each of the two parts includes a
chapter of necessary reviews, two main chapters and a chapter summarizing other
works related to the concerned part. Part 1 includes a first chapter containing
an introduction and some necessary reviews; two main chapters, one on the
max-plus algebra model for the train dynamics on a metro line, the other one on
the network calculus approach for modeling and calculating performance bounds
on road networks; and a final chapter summarizing my other contributions on the
topic of this part. Part 2 includes a first chapter containing an introduction
and some necessary reviews; two main chapters, one on the microscopic modeling
of traffic taking into account anticipation in driving, the other one on the
modeling of the train dynamics on a metro line taking into account the
passenger travel demand; and a final chapter summarizing my other contributions
on the topic of this part.
| arxiv topic:math.OC math.DS |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.