id stringlengths 16 29 | text stringlengths 86 3.49k | source stringlengths 14 112 |
|---|---|---|
arxiv_dataset-87001707.04968 | Visual Question Answering with Memory-Augmented Networks
cs.CV cs.CL
In this paper, we exploit a memory-augmented neural network to predict
accurate answers to visual questions, even when those answers occur rarely in
the training set. The memory network incorporates both internal and external
memory blocks and selectively pays attention to each training exemplar. We show
that memory-augmented neural networks are able to maintain a relatively
long-term memory of scarce training exemplars, which is important for visual
question answering due to the heavy-tailed distribution of answers in a general
VQA setting. Experimental results on two large-scale benchmark datasets show
the favorable performance of the proposed algorithm with a comparison to state
of the art.
| arxiv topic:cs.CV cs.CL |
arxiv_dataset-87011707.05068 | The cosmic QCD phase transition with dense matter and its gravitational
waves from holography
hep-th gr-qc hep-ph
Consistent with cosmological constraints, there are scenarios with the large
lepton asymmetry which can lead to the finite baryochemical potential at the
cosmic QCD phase transition scale. In this paper, we investigate this
possibility in the holographic models. Using the holographic renormalization
method, we find the first order Hawking-Page phase transition, between
Reissner-Nordstr$\rm\ddot{o}$m AdS black hole and thermal charged AdS space,
corresponding to the de/confinement phase transition. We obtain the
gravitational wave spectra generated during the evolution of bubbles for a
range of the bubble wall velocity and examine the reliability of the scenarios
and consequent calculations by gravitational wave experiments.
| arxiv topic:hep-th gr-qc hep-ph |
arxiv_dataset-87021707.05168 | Use of ANTARES and IceCube data to constrain a single power-law neutrino
flux
hep-ph astro-ph.HE hep-ex
We perform the first statistical combined analysis of the diffuse neutrino
flux observed by ANTARES (nine-year) and IceCube (six-year) by assuming a
single astrophysical power-law flux. The combined analysis reduces by a few
percent the best-fit values for the flux normalization and the spectral index.
Both data samples show an excess in the same energy range (40--200 TeV),
suggesting the presence of a second component. We perform a goodness-of-fit
test to scrutinize the null assumption of a single power-law, scanning
different values for the spectral index. The addition of the ANTARES data
reduces the $p$-value by a factor 2$\div$3. In particular, a single power-law
component in the neutrino flux with the spectral index deduced by the six-year
up-going muon neutrinos of IceCube is disfavored with a $p$-value smaller than
$10^{-2}$.
| arxiv topic:hep-ph astro-ph.HE hep-ex |
arxiv_dataset-87031707.05268 | A new veto for continuous gravitational wave searches
gr-qc astro-ph.IM
We present a new veto procedure to distinguish between continuous
gravitational wave (CW) signals and the detector artifacts that can mimic their
behavior. The veto procedure exploits the fact that a long-lasting coherent
disturbance is less likely than a real signal to exhibit a Doppler modulation
of astrophysical origin. Therefore, in the presence of an outlier from a
search, we perform a multi-step search around the frequency of the outlier with
the Doppler modulation turned off (DM-off), and compare these results with the
results from the original (DM-on) search. If the results from the DM-off search
are more significant than those from the DM-on search, the outlier is most
likely due to an artifact rather than a signal. We tune the veto procedure so
that it has a very low false dismissal rate. With this veto, we are able to
identify as coherent disturbances >99.9% of the 6349 candidates from the recent
all-sky low-frequency Einstein@Home search on the data from the Advanced LIGO
O1 observing run [1]. We present the details of each identified disturbance in
the Appendix.
| arxiv topic:gr-qc astro-ph.IM |
arxiv_dataset-87041707.05368 | A robotic vision system to measure tree traits
cs.RO
The autonomous measurement of tree traits, such as branching structure,
branch diameters, branch lengths, and branch angles, is required for tasks such
as robotic pruning of trees as well as structural phenotyping. We propose a
robotic vision system called the Robotic System for Tree Shape Estimation
(RoTSE) to determine tree traits in field settings. The process is composed of
the following stages: image acquisition with a mobile robot unit, segmentation,
reconstruction, curve skeletonization, conversion to a graph representation,
and then computation of traits. Quantitative and qualitative results on apple
trees are shown in terms of accuracy, computation time, and robustness.
Compared to ground truth measurements, the RoTSE produced the following
estimates: branch diameter (root mean-squared error $2.97$ mm), branch length
(root mean-squared error $136.92$ mm), and branch angle (mean-squared error
$31.07$ degrees). The average run time was $8.47$ minutes when the voxel
resolution was $3$ mm$^3$.
| arxiv topic:cs.RO |
arxiv_dataset-87051707.05468 | Detecting Intentional Lexical Ambiguity in English Puns
cs.CL
The article describes a model of automatic analysis of puns, where a word is
intentionally used in two meanings at the same time (the target word). We
employ Roget's Thesaurus to discover two groups of words which, in a pun, form
around two abstract bits of meaning (semes). They become a semantic vector,
based on which an SVM classifier learns to recognize puns, reaching a score
0.73 for F-measure. We apply several rule-based methods to locate intentionally
ambiguous (target) words, based on structural and semantic criteria. It appears
that the structural criterion is more effective, although it possibly
characterizes only the tested dataset. The results we get correlate with the
results of other teams at SemEval-2017 competition (Task 7 Detection and
Interpretation of English Puns) considering effects of using supervised
learning models and word statistics.
| arxiv topic:cs.CL |
arxiv_dataset-87061707.05568 | Solitary waves and double layers in an adiabatic multi-component space
plasma
physics.plasm-ph
The formation and propagation of small amplitude Heavy-ion-acoustic (HIA)
solitary waves and double layers in an unmagnetized collisionless
multi-component plasma system consisting of superthermal electrons, Boltzmann
distributed light ions, and adiabatic positively charged inertial heavy ions
are theoretically investigated. The reductive perturbation technique is
employed to derive the modified Korteweg-de Vries (mK-dV) and standard Gardner
(SG) equations. The solitary wave (SW) solution of mK-dV and SG equations as
well as Double Layers (DLs) solution of SG equation is studied for analysis of
higher-order nonlinearity. It is found that the plasma system under
consideration supports positive and negative potential Gardner solitons but
only positive potential mK-dV solitons. In addition, it is shown that, the
basic properties of HIA mK-dV and Gardner solitons and DLs (viz. polarity,
amplitude, width, and phase speed) are incomparably influenced by the
adiabaticity effect of heavy ions and the superthermality effect of electrons.
The relevance of the present findings to the system of space plasmas as well as
to the system of researchers interest is specified.
| arxiv topic:physics.plasm-ph |
arxiv_dataset-87071707.05668 | Empirical evaluation of a Q-Learning Algorithm for Model-free Autonomous
Soaring
cs.LG
Autonomous unpowered flight is a challenge for control and guidance systems:
all the energy the aircraft might use during flight has to be harvested
directly from the atmosphere. We investigate the design of an algorithm that
optimizes the closed-loop control of a glider's bank and sideslip angles, while
flying in the lower convective layer of the atmosphere in order to increase its
mission endurance. Using a Reinforcement Learning approach, we demonstrate the
possibility for real-time adaptation of the glider's behaviour to the
time-varying and noisy conditions associated with thermal soaring flight. Our
approach is online, data-based and model-free, hence avoids the pitfalls of
aerological and aircraft modelling and allow us to deal with uncertainties and
non-stationarity. Additionally, we put a particular emphasis on keeping low
computational requirements in order to make on-board execution feasible. This
article presents the stochastic, time-dependent aerological model used for
simulation, together with a standard aircraft model. Then we introduce an
adaptation of a Q-learning algorithm and demonstrate its ability to control the
aircraft and improve its endurance by exploiting updrafts in non-stationary
scenarios.
| arxiv topic:cs.LG |
arxiv_dataset-87081707.05768 | Ask Me Anything: A Conversational Interface to Augment Information
Security Workers
cs.HC
Security products often create more problems than they solve, drowning users
in alerts without providing the context required to remediate threats. This
challenge is compounded by a lack of experienced personnel and security tools
with complex interfaces. These interfaces require users to become domain
experts or rely on repetitive, time consuming tasks to turn this data deluge
into actionable intelligence. In this paper we present Artemis, a
conversational interface to endpoint detection and response (EDR) event data.
Artemis leverages dialog to drive the automation of complex tasks and reduce
the need to learn a structured query language. Designed to empower
inexperienced and junior security workers to better understand their security
environment, Artemis provides an intuitive platform to ask questions of alert
data as users are guided through triage and hunt workflows. In this paper, we
will discuss our user-centric design methodology, feedback from user
interviews, and the design requirements generated upon completion of our study.
We will also present core functionality, findings from scenario-based testing,
and future research for the Artemis platform.
| arxiv topic:cs.HC |
arxiv_dataset-87091707.05868 | From anti-perovskite to double anti-perovskite: lattice chemistry basis
for super-fast transportation of Li+ ions in cubic solid lithium
halogen-chalcogenides
physics.chem-ph cond-mat.mtrl-sci
Using a materials genome approach on the basis of the density functional
theory, we have formulated a new class of inorganic electrolytes for fast
diffusion of Li+ ions, through fine-tuning of lattice chemistry of
anti-perovskite structures. Systematic modelling has been carried out to
elaborate the structural stability and ion transportation characteristics in
Li3AX based cubic anti-perovskite, through alloying on the chalcogen lattice
site (A) and alternative occupancy of the halogen site (X). In addition to
identifying effective ways for reduction of diffusion barriers for Li+ ions in
anti-perovskite phases via suitable designation of lattice occupancy, the
current theoretical study leads to discovery and synthesis of a new phase with
a double-anti-perovskite structure, Li6OSI2 (or Li3O0.5S0.5I). Such a new
compound is of fairly low activation barrier for Li+ diffusion, together with a
wide energy band gap to hinder conduction of electrons.
| arxiv topic:physics.chem-ph cond-mat.mtrl-sci |
arxiv_dataset-87101707.05968 | Parameterized complexity of games with monotonically ordered
{\omega}-regular objectives
cs.GT cs.LO
In recent years, two-player zero-sum games with multiple objectives have
received a lot of interest as a model for the synthesis of complex reactive
systems. In this framework, Player 1 wins if he can ensure that all objectives
are satisfied against any behavior of Player 2. When this is not possible to
satisfy all the objectives at once, an alternative is to use some preorder on
the objectives according to which subset of objectives Player 1 wants to
satisfy. For example, it is often natural to provide more significance to one
objective over another, a situation that can be modelled with lexicographically
ordered objectives for instance. Inspired by recent work on concurrent games
with multiple {\omega}-regular objectives by Bouyer et al., we investigate in
detail turned-based games with monotonically ordered and {\omega}-regular
objectives. We study the threshold problem which asks whether player 1 can
ensure a payoff greater than or equal to a given threshold w.r.t. a given
monotonic preorder. As the number of objectives is usually much smaller than
the size of the game graph, we provide a parametric complexity analysis and we
show that our threshold problem is in FPT for all monotonic preorders and all
classical types of {\omega}-regular objectives. We also provide polynomial time
algorithms for B\"uchi, coB\"uchi and explicit Muller objectives for a large
subclass of monotonic preorders that includes among others the lexicographic
preorder. In the particular case of lexicographic preorder, we also study the
complexity of computing the values and the memory requirements of optimal
strategies.
| arxiv topic:cs.GT cs.LO |
arxiv_dataset-87111707.06068 | On Finding Maximum Cardinality Subset of Vectors with a Constraint on
Normalized Squared Length of Vectors Sum
cs.DS cs.CV
In this paper, we consider the problem of finding a maximum cardinality
subset of vectors, given a constraint on the normalized squared length of
vectors sum. This problem is closely related to Problem 1 from (Eremeev,
Kel'manov, Pyatkin, 2016). The main difference consists in swapping the
constraint with the optimization criterion.
We prove that the problem is NP-hard even in terms of finding a feasible
solution. An exact algorithm for solving this problem is proposed. The
algorithm has a pseudo-polynomial time complexity in the special case of the
problem, where the dimension of the space is bounded from above by a constant
and the input data are integer. A computational experiment is carried out,
where the proposed algorithm is compared to COINBONMIN solver, applied to a
mixed integer quadratic programming formulation of the problem. The results of
the experiment indicate superiority of the proposed algorithm when the
dimension of Euclidean space is low, while the COINBONMIN has an advantage for
larger dimensions.
| arxiv topic:cs.DS cs.CV |
arxiv_dataset-87121707.06168 | Channel Pruning for Accelerating Very Deep Neural Networks
cs.CV
In this paper, we introduce a new channel pruning method to accelerate very
deep convolutional neural networks.Given a trained CNN model, we propose an
iterative two-step algorithm to effectively prune each layer, by a LASSO
regression based channel selection and least square reconstruction. We further
generalize this algorithm to multi-layer and multi-branch cases. Our method
reduces the accumulated error and enhance the compatibility with various
architectures. Our pruned VGG-16 achieves the state-of-the-art results by 5x
speed-up along with only 0.3% increase of error. More importantly, our method
is able to accelerate modern networks like ResNet, Xception and suffers only
1.4%, 1.0% accuracy loss under 2x speed-up respectively, which is significant.
Code has been made publicly available.
| arxiv topic:cs.CV |
arxiv_dataset-87131707.06268 | On Newstead's Mayer-Vietoris argument in characteristic 2
math.GT
Consider the moduli space of framed flat $U(2)$ connections with fixed odd
determinant over a surface. Newstead combined some fundamental facts about this
moduli space with the Mayer-Vietoris sequence to compute its betti numbers over
any field not of characteristic two. We adapt his method in characteristic two
to produce conjectural recursive formulae for the mod two betti numbers of the
framed moduli space which we partially verify. We also discuss the interplay
with the mod two cohomology ring structure of the unframed moduli space.
| arxiv topic:math.GT |
arxiv_dataset-87141707.06368 | Some properties for the Steklov averages
math.AP
We derive and present a collection of properties about the Steklov averages,
including some results about the derivation with respect to spatial variables,
and with respect to time, and a form of the fundamental theorem of the
calculus.
| arxiv topic:math.AP |
arxiv_dataset-87151707.06468 | Breaking the Nonsmooth Barrier: A Scalable Parallel Method for Composite
Optimization
math.OC cs.LG stat.ML
Due to their simplicity and excellent performance, parallel asynchronous
variants of stochastic gradient descent have become popular methods to solve a
wide range of large-scale optimization problems on multi-core architectures.
Yet, despite their practical success, support for nonsmooth objectives is still
lacking, making them unsuitable for many problems of interest in machine
learning, such as the Lasso, group Lasso or empirical risk minimization with
convex constraints.
In this work, we propose and analyze ProxASAGA, a fully asynchronous sparse
method inspired by SAGA, a variance reduced incremental gradient algorithm. The
proposed method is easy to implement and significantly outperforms the state of
the art on several nonsmooth, large-scale problems. We prove that our method
achieves a theoretical linear speedup with respect to the sequential version
under assumptions on the sparsity of gradients and block-separability of the
proximal term. Empirical benchmarks on a multi-core architecture illustrate
practical speedups of up to 12x on a 20-core machine.
| arxiv topic:math.OC cs.LG stat.ML |
arxiv_dataset-87161707.06568 | UV Exposed Optical Fibers with Frequency Domain Reflectometry for Device
Tracking in Intra-Arterial Procedures
physics.med-ph
Shape tracking of medical devices using strain sensing properties in optical
fibers has seen increased attention in recent years. In this paper, we propose
a novel guidance system for intra-arterial procedures using a distributed
strain sensing device based on optical frequency domain reflectometry (OFDR) to
track the shape of a catheter. Tracking enhancement is provided by exposing a
fiber triplet to a focused ultraviolet beam, producing high scattering
properties. Contrary to typical quasi-distributed strain sensors, we propose a
truly distributed strain sensing approach, which allows to reconstruct a fiber
triplet in real-time. A 3D roadmap of the hepatic anatomy integrated with a 4D
MR imaging sequence allows to navigate the catheter within the
pre-interventional anatomy, and map the blood flow velocities in the arterial
tree. We employed Riemannian anisotropic heat kernels to map the sensed data to
the pre-interventional model. Experiments in synthetic phantoms and an in vivo
model are presented. Results show that the tracking accuracy is suitable for
interventional tracking applications, with a mean 3D shape reconstruction
errors of 1.6 +/- 0.3 mm. This study demonstrates the promising potential of
MR-compatible UV-exposed OFDR optical fibers for non-ionizing device guidance
in intra-arterial procedures.
| arxiv topic:physics.med-ph |
arxiv_dataset-87171707.06668 | Are Some Technologies Beyond Regulatory Regimes?
physics.soc-ph
Regulatory frameworks are a common tool in governance to incent and coerce
behaviors supporting national or strategic stability. This includes domestic
regulations and international agreements. Though regulation is always a
challenge, the domain of fast evolving threats, like cyber, are proving much
more difficult to control. Many discussions are underway searching for
approaches that can provide national security in these domains. We use game
theoretic learning models to explore the question of strategic stability with
respect to the democratization of certain technologies (such as cyber). We
suggest that such many-player games could inherently be chaotic with no
corresponding (Nash) equilibria. In the absence of such equilibria, traditional
approaches, as measures to achieve levels of overall security, may not be
suitable approaches to support strategic stability in these domains. Altogether
new paradigms may be needed for these issues. At the very least, regulatory
regimes that fail to address the basic nature of the technology domains should
not be pursued as a default solution, regardless of success in other domains.
In addition, the very chaotic nature of these domains may hold the promise of
novel approaches to regulation.
| arxiv topic:physics.soc-ph |
arxiv_dataset-87181707.06768 | Integrability conditions for Compound Random Measures
stat.ME math.PR math.ST stat.TH
Compound random measures (CoRM's) are a flexible and tractable framework for
vectors of completely random measure. In this paper, we provide conditions to
guarantee the existence of a CoRM. Furthermore, we prove some interesting
properties of CoRM's when exponential scores and regularly varying L\'evy
intensities are considered.
| arxiv topic:stat.ME math.PR math.ST stat.TH |
arxiv_dataset-87191707.06868 | Nilpotency and strong nilpotency for finite semigroups
math.GR
Nilpotent semigroups in the sense of Mal'cev are defined by semigroup
identities. Finite nilpotent semigroups constitute a pseudovariety,
$\mathsf{MN}$, which has finite rank. The semigroup identities that define
nilpotent semigroups, lead us to define strongly Mal'cev nilpotent semigroups.
Finite strongly Mal'cev nilpotent semigroups constitute a non-finite rank
pseudovariety, $\mathsf{SMN}$. The pseudovariety $\mathsf{SMN}$ is strictly
contained in the pseudovariety $\mathsf{MN}$ but all finite nilpotent groups
are in $\mathsf{SMN}$. We show that the pseudovariety $\mathsf{MN}$ is the
intersection of the pseudovariety $\mathsf{BG_{nil}}$ with a pseudovariety
defined by a $\kappa$-identity. We further compare the pseudovarieties
$\mathsf{MN}$ and $\mathsf{SMN}$ with the Mal'cev product of the
pseudovarieties $\mathsf{J}$ and $\mathsf{G_{nil}}$.
| arxiv topic:math.GR |
arxiv_dataset-87201707.06968 | Modelling wormholes in $f(R,T)$ gravity
gr-qc
In this work we propose the modelling of static wormholes within the $f(R,T)$
extended theory of gravity perspective. We present some models of wormholes,
which are constructed from different hypothesis for their matter content, i.e.,
different relations for their pressure components (radial and lateral) and
different equations of state. The solutions obtained for the shape function of
the wormholes obey the necessary metric conditions. They show a behaviour
similar to those found in previous references about wormholes, which also
happens to our solutions for the energy density of such objects. We also apply
the energy conditions for the wormholes physical content.
| arxiv topic:gr-qc |
arxiv_dataset-87211707.07068 | Adaptive Channel Prediction, Beamforming and Scheduling Design for 5G
V2I Network
cs.IT math.IT
One of the important use-cases of 5G network is the vehicle to infrastructure
(V2I) communication which requires accurate understanding about its dynamic
propagation environment. As 5G base stations (BSs) tend to have multiple
antennas, they will likely employ beamforming to steer their radiation pattern
to the desired vehicle equipment (VE). Furthermore, since most wireless
standards employ an OFDM system, each VE may use one or more sub-carriers. To
this end, this paper proposes a joint design of adaptive channel prediction,
beamforming and scheduling for 5G V2I communications. The channel prediction
algorithm is designed without the training signal and channel impulse response
(CIR) model. In this regard, first we utilize the well known adaptive recursive
least squares (RLS) technique for predicting the next block CIR from the past
and current block received signals (a block may have one or more OFDM symbols).
Then, we jointly design the beamforming and VE scheduling for each sub-carrier
to maximize the uplink channel average sum rate by utilizing the predicted CIR.
The beamforming problem is formulated as a Rayleigh quotient optimization where
its global optimal solution is guaranteed. And, the VE scheduling design is
formulated as an integer programming problem which is solved by employing a
greedy search. The superiority of the proposed channel prediction and
scheduling algorithms over those of the existing ones is demonstrated via
numerical simulations.
| arxiv topic:cs.IT math.IT |
arxiv_dataset-87221707.07168 | Order by quenched disorder in the model triangular antiferromagnet
RbFe(MoO4)2
cond-mat.str-el
We observe a disappearance of the 1/3 magnetization plateau and a striking
change of the magnetic configuration under a moderate doping of the model
triangular antiferromagnet RbFe(MoO4)2. The reason is an effective lifting of
degeneracy of mean-field ground states by a random potential of impurities,
which compensates, in the low temperature limit, the fluctuation contribution
to free energy. These results provide a direct experimental confirmation of the
fluctuation origin of the ground state in a real frustrated system. The change
of the ground state to a least collinear configuration reveals an effective
positive biquadratic exchange provided by the structural disorder. On heating,
doped samples regain the structure of a pure compound thus allowing for an
investigation of the remarkable competition between thermal and structural
disorder.
| arxiv topic:cond-mat.str-el |
arxiv_dataset-87231707.07268 | $^1S_0$ pairing in neutron matter
nucl-th cond-mat.quant-gas
We report calculations of the superfluid pairing gap in neutron matter for
the $^1S_0$ components of the Reid soft-core $V_6$ and the Argonne $V_{4}'$
two-nucleon interactions. Ground-state calculations have been carried out using
the central part of the operator-basis representation of these interactions to
determine optimal Jastrow-Feenberg correlations and corresponding effective
pairing interactions within the correlated-basis formalism (CBF), the required
matrix elements in the correlated basis being evaluated by Fermi
hypernetted-chain techniques. Different implementations of the
Fermi-Hypernetted Chain Euler-Lagrange method (FHNC-EL) agree at the percent
level up to nuclear matter saturation density. For the assumed interactions,
which are realistic within the low density range involved in $^1S_0$ neutron
pairing, we did not find a dimerization instability arising from divergence of
the in-medium scattering length, as was reported recently for simple
square-well and Lennard-Jones potential models (Phys. Rev. A {\bf 92}, 023640
(2015)).
| arxiv topic:nucl-th cond-mat.quant-gas |
arxiv_dataset-87241707.07368 | Common Denominator for Value and Expectation No-Go Theorems
quant-ph
Hidden-variable (HV) theories allege that a quantum state describes an
ensemble of systems distinguished by the values of hidden variables. No-go
theorems assert that HV theories cannot match the predictions of quantum
theory. The present work started with repairing flaws in the literature on
no-go theorems asserting that HV theories cannot predict the expectation values
of measurements. That literature gives one an impression that expectation no-go
theorems subsume the time-honored no-go theorems asserting that HV theories
cannot predict the possible values of measurements. But the two approaches
speak about different kinds of measurement. This hinders comparing them to each
other. Only projection measurements are common to both. Here, we sharpen the
results of both approaches so that only projection measurements are used. This
allows us to clarify the similarities and differences between the two
approaches. Neither one dominates the other.
| arxiv topic:quant-ph |
arxiv_dataset-87251707.07468 | Finite presheaves and $A$-finite generation of unstable algebras mod
nilpotents
math.AT
Inspired by the work of Henn, Lannes and Schwartz on unstable algebras over
the Steenrod algebra modulo nilpotents, a characterization of unstable algebras
that are $A$-finitely generated up to nilpotents is given in terms of the
associated presheaf, by introducing the notion of a finite presheaf. In
particular, this gives the natural characterization of the (co)analytic
presheaves that are important in the theory of Henn, Lannes and Schwartz.
However, finite presheaves remain imperfectly understood, as illustrated by
examples. One important class of examples is shown to be provided by unstable
algebras of finite transcendence degree (under a necessary weak finiteness
condition).
For unstable Hopf algebras, it is shown that the situation is much better:
the associated presheaf is finite if and only if its growth function is
polynomial. This leads to a description of unstable Hopf algebras modulo
nilpotents in the spirit of Henn, Lannes and Schwartz.
| arxiv topic:math.AT |
arxiv_dataset-87261707.07568 | CAp 2017 challenge: Twitter Named Entity Recognition
cs.CL
The paper describes the CAp 2017 challenge. The challenge concerns the
problem of Named Entity Recognition (NER) for tweets written in French. We
first present the data preparation steps we followed for constructing the
dataset released in the framework of the challenge. We begin by demonstrating
why NER for tweets is a challenging problem especially when the number of
entities increases. We detail the annotation process and the necessary
decisions we made. We provide statistics on the inter-annotator agreement, and
we conclude the data description part with examples and statistics for the
data. We, then, describe the participation in the challenge, where 8 teams
participated, with a focus on the methods employed by the challenge
participants and the scores achieved in terms of F$_1$ measure. Importantly,
the constructed dataset comprising $\sim$6,000 tweets annotated for 13 types of
entities, which to the best of our knowledge is the first such dataset in
French, is publicly available at \url{http://cap2017.imag.fr/competition.html} .
| arxiv topic:cs.CL |
arxiv_dataset-87271707.07668 | Unlocking Sensitivity for Visibility-based Estimators of the 21 cm
Reionization Power Spectrum
astro-ph.IM astro-ph.CO
Radio interferometers designed to measure the cosmological 21 cm power
spectrum require high sensitivity. Several modern low-frequency interferometers
feature drift-scan antennas placed on a regular grid to maximize the number of
instantaneously coherent (redundant) measurements. However, even for such
maximum-redundancy arrays, significant sensitivity comes through partial
coherence between baselines. Current visibility-based power-spectrum pipelines,
though shown to ease control of systematics, lack the ability to make use of
this partial redundancy. We introduce a method to leverage partial redundancy
in such power-spectrum pipelines for drift-scan arrays. Our method
cross-multiplies baseline pairs at a time lag and quantifies the sensitivity
contributions of each pair of baselines. Using the configurations and beams of
the 128-element Donald C. Backer Precision Array for Probing the Epoch of
Reionization (PAPER-128) and staged deployments of the Hydrogen Epoch of
Reionization Array, we illustrate how our method applies to different arrays
and predict the sensitivity improvements associated with pairing partially
coherent baselines. As the number of antennas increases, we find partial
redundancy to be of increasing importance in unlocking the full sensitivity of
upcoming arrays.
| arxiv topic:astro-ph.IM astro-ph.CO |
arxiv_dataset-87281707.07768 | Morphometric analysis of polygonal cracking patterns in desiccated
starch slurries
cond-mat.soft cond-mat.dis-nn cond-mat.mtrl-sci cond-mat.stat-mech nlin.PS
We investigate the geometry of two-dimensional polygonal cracking that forms
on the air-exposed surface of dried starch slurries. Two different kinds of
starches, made from potato and corn, exhibited distinguished crack evolution,
and there were contrasting effects of slurry thickness on the probability
distribution of the polygonal cell area. The experimental findings are believed
to result from the difference in the shape and size of starch grains, which
strongly influence the capillary transport of water and tensile stress field
that drives the polygonal cracking.
| arxiv topic:cond-mat.soft cond-mat.dis-nn cond-mat.mtrl-sci cond-mat.stat-mech nlin.PS |
arxiv_dataset-87291707.07868 | Projective structures, neighborhoods of rational curves and Painlev'e
equations
math.CA math.CV math.DG
We investigate the duality between local (complex analytic) projective
structures on surfaces and two dimensional (complex analytic) neighborhoods of
rational curves having self-intersection +1. We study the analytic
classification, existence of normal forms, pencil/fibration decomposition,
infinitesimal symmetries. We deduce some transcendental result about Painlev'e
equations.
| arxiv topic:math.CA math.CV math.DG |
arxiv_dataset-87301707.07968 | Classical conformal blocks and isomonodromic deformations
hep-th
The leading classical asymptotics of Virasoro conformal blocks on the Riemann
sphere with n generic and n-3 "heavy" degenerate field insertions can be
described in terms of the geometry of Garnier system describing the monodromy
preserving deformations of second order Fuchsian differential equations on an
n-punctured sphere. This allows us to characterise the leading classical
asymptotics of Virasoro conformal blocks completely, and to clarify in which
sense conformal field theory represents a quantisation of the isomonodromic
deformation problem.
| arxiv topic:hep-th |
arxiv_dataset-87311707.08068 | Weak vorticity formulation of 2D Euler equations with white noise
initial condition
math.PR
The 2D Euler equations with random initial condition distributed as a certain
Gaussian measure are considered. The theory developed by S. Albeverio and A.-B.
Cruzeiro is revisited, following the approach of weak vorticity formulation. A
solution is constructed as a limit of random point vortices. This allows to
prove that it is also limit of L^\infty-vorticity solutions. The result is
generalized to initial measures that have a continuous bounded density with
respect to the original Gaussian measure.
| arxiv topic:math.PR |
arxiv_dataset-87321707.08168 | Estimating proton beam energy spread using Bragg peak measurement
physics.med-ph
230 MeV proton beam out of a cyclotron was delivered into a Zebra multi
layered IC detector (IBA) calibrated in terms of penetration range in water.
The analysis of the measured Bragg peak determines penetration range in water
which can be subsequently converted into proton beam energy using Range-Energy
tables. We extended this analysis to obtain an estimate of the beam energy
spread out of the cyclotron. Using Monte Carlo simulations we established the
correlation between Bragg peak shape parameters (width at 50% and 80% dose
levels, distal falloff) and penetration range for a monoenergetic proton beam.
Then we studied how this correlation changes when the shape of Bragg peak is
distorted by the beam focusing conditions. We found that small field size or
diverging beam cause Bragg peak deformation predominantly in the proximal
region. The distal shape of the renormalized Bragg peaks stays nearly constant.
This excludes usage of Bragg peak width parameters for energy spread estimates.
The measured Bragg peaks had an average distal falloff of 4.86mm, which
corresponds to an effective range of 35.5cm for a monoenergetic beam. The
32.7cm measured penetration range is 2.8cm less. Passage of a 230 MeV proton
beam through a 2.8cm thick slab of water results in a 0.56 MeV energy spread.
As a final check, we confirmed agreement between shapes of the measured Bragg
peak and one generated by Monte-Carlo code for proton beam with 0.56 MeV energy
spread.
| arxiv topic:physics.med-ph |
arxiv_dataset-87331707.08268 | Incidence Results and Bounds of Trilinear and Quadrilinear Exponential
Sums
math.NT
We give a new bound on the number of collinear triples for two arbitrary
subsets of a finite field. This improves on existing results which rely on the
Cauchy inequality. We then us this to provide a new bound on trilinear and
quadrilinear exponential sums.
| arxiv topic:math.NT |
arxiv_dataset-87341707.08368 | Quasiconvex elastodynamics: weak-strong uniqueness for measure-valued
solutions
math.AP
A weak-strong uniqueness result is proved for measure-valued solutions to the
system of conservation laws arising in elastodynamics. The main novelty brought
forward by the present work is that the underlying stored-energy function of
the material is assumed strongly quasiconvex. The proof employs tools from the
calculus of variations to establish general convexity-type bounds on
quasiconvex functions and recasts them in order to adapt the relative entropy
method to quasiconvex elastodynamics.
| arxiv topic:math.AP |
arxiv_dataset-87351707.08468 | A Decidable Very Expressive Description Logic for Databases (Extended
Version)
cs.AI
We introduce $\mathcal{DLR}^+$, an extension of the n-ary propositionally
closed description logic $\mathcal{DLR}$ to deal with attribute-labelled tuples
(generalising the positional notation), projections of relations, and global
and local objectification of relations, able to express inclusion, functional,
key, and external uniqueness dependencies. The logic is equipped with both TBox
and ABox axioms. We show how a simple syntactic restriction on the appearance
of projections sharing common attributes in a $\mathcal{DLR}^+$ knowledge base
makes reasoning in the language decidable with the same computational
complexity as $\mathcal{DLR}$. The obtained $\mathcal{DLR}^\pm$ n-ary
description logic is able to encode more thoroughly conceptual data models such
as EER, UML, and ORM.
| arxiv topic:cs.AI |
arxiv_dataset-87361707.08568 | Axion as a cold dark matter candidate: Proof to fully nonlinear order
gr-qc hep-ph hep-th
We present a proof of the axion as a cold dark matter candidate to the fully
nonlinear order perturbations based on Einstein's gravity. We consider the
axion as a coherently oscillating massive classical scalar field without
interaction. We present the fully nonlinear and exact, except for {\it
ignoring} the transverse-tracefree tensor-type perturbation, hydrodynamic
equations for an axion fluid in Einstein's gravity. We show that the axion has
the characteristic pressure and anisotropic stress, the latter starts to appear
from the second-order perturbation. But these terms do not directly affect the
hydrodynamic equations in our axion treatment. Instead, what behaves as the
effective pressure term in relativistic hydrodynamic equations is the perturbed
lapse function and the relativistic result coincides exactly with the one known
in the previous non-relativistic studies. The effective pressure term leads to
a Jeans scale which is of the solar-system scale for conventional axion mass.
As the fully nonlinear and relativistic hydrodynamic equations for an axion
fluid coincide exactly with the ones of a zero-pressure fluid in the
super-Jeans scale, we have proved the cold dark matter nature of such an axion
in that scale.
| arxiv topic:gr-qc hep-ph hep-th |
arxiv_dataset-87371707.08668 | A Tale of Two DRAGGNs: A Hybrid Approach for Interpreting
Action-Oriented and Goal-Oriented Instructions
cs.AI cs.CL
Robots operating alongside humans in diverse, stochastic environments must be
able to accurately interpret natural language commands. These instructions
often fall into one of two categories: those that specify a goal condition or
target state, and those that specify explicit actions, or how to perform a
given task. Recent approaches have used reward functions as a semantic
representation of goal-based commands, which allows for the use of a
state-of-the-art planner to find a policy for the given task. However, these
reward functions cannot be directly used to represent action-oriented commands.
We introduce a new hybrid approach, the Deep Recurrent Action-Goal Grounding
Network (DRAGGN), for task grounding and execution that handles natural
language from either category as input, and generalizes to unseen environments.
Our robot-simulation results demonstrate that a system successfully
interpreting both goal-oriented and action-oriented task specifications brings
us closer to robust natural language understanding for human-robot interaction.
| arxiv topic:cs.AI cs.CL |
arxiv_dataset-87381707.08768 | Equivariant extensions of Ga-torsors over punctured surfaces
math.AG
Motivated by the study of the structure of algebraic actions the additive
group on affine threefolds X, we consider a special class of such varieties
whose algebraic quotient morphisms X $\rightarrow$ X//Ga restrict to principal
homogeneous bundles over the complement of a smooth point of the quotient. We
establish basic general properties of these varieties and construct families of
examples illustrating their rich geometry. In particular, we give a complete
classification of a natural subclass consisting of threefolds X endowed with
proper Ga-actions, whose algebraic quotient morphisms $\pi$ : X $\rightarrow$
X//Ga are surjective with only isolated degenerate fibers, all isomorphic to
the affine plane A 2 when equipped with their reduced structures.
| arxiv topic:math.AG |
arxiv_dataset-87391707.08868 | Importance sampling for metastable and multiscale dynamical systems
math.PR math.OC stat.ME
In this article, we address the issues that come up in the design of
importance sampling schemes for rare events associated to stochastic dynamical
systems. We focus on the issue of metastability and on the effect of multiple
scales. We discuss why seemingly reasonable schemes that follow large
deviations optimal paths may perform poorly in practice, even though they are
asymptotically optimal. Pre-asymptotic optimality is important when one deals
with metastable dynamics and we discuss possible ways as to how to address this
issue. Moreover, we discuss how the effect of the multiple scales (either in
periodic or random environments) on the efficient design of importance sampling
should be addressed. We discuss the mathematical and practical issues that come
up, how to overcome some of the issues and discuss future challenges.
| arxiv topic:math.PR math.OC stat.ME |
arxiv_dataset-87401707.08968 | Resilience of hidden order to symmetry-preserving disorder
cond-mat.stat-mech cond-mat.quant-gas quant-ph
We study the robustness of non-local string order in two paradigmatic
disordered spin-chain models, a spin-1/2 cluster-Ising and a spin-1 XXZ
Heisenberg chain. In the clean case, they both display a transition from
antiferromagnetic to string order. Applying a disorder which preserves the
Hamiltonian symmetries, we find that the transition persists in both models. In
the disordered cluster-Ising model we can study the transition analytically --
by applying the strongest coupling renormalization group -- and numerically --
by exploiting integrability to study the antiferromagnetic and string order
parameters. We map the model into a quadratic fermion chain, where the
transition appears as a change in the number of zero-energy edge modes. We also
explore its zero-temperature-singularity behavior and find a transition from a
non-singular to a singular region, at a point that is different from the one
separating non-local and local ordering.} The disordered Heisenberg chain can
be treated only numerically: by means of MPS-based simulations, we are able to
locate the existence of a transition between antiferromagnetic and
string-ordered phase, through the study of order parameters. Finally we discuss
possible connections of our findings with many body localization.
| arxiv topic:cond-mat.stat-mech cond-mat.quant-gas quant-ph |
arxiv_dataset-87411707.09068 | Tartan: Accelerating Fully-Connected and Convolutional Layers in Deep
Learning Networks by Exploiting Numerical Precision Variability
cs.NE
Tartan (TRT), a hardware accelerator for inference with Deep Neural Networks
(DNNs), is presented and evaluated on Convolutional Neural Networks. TRT
exploits the variable per layer precision requirements of DNNs to deliver
execution time that is proportional to the precision p in bits used per layer
for convolutional and fully-connected layers. Prior art has demonstrated an
accelerator with the same execution performance only for convolutional layers.
Experiments on image classification CNNs show that on average across all
networks studied, TRT outperforms a state-of-the-art bit-parallel accelerator
by 1:90x without any loss in accuracy while it is 1:17x more energy efficient.
TRT requires no network retraining while it enables trading off accuracy for
additional improvements in execution performance and energy efficiency. For
example, if a 1% relative loss in accuracy is acceptable, TRT is on average
2:04x faster and 1:25x more energy efficient than a conventional bit-parallel
accelerator. A Tartan configuration that processes 2-bits at time, requires
less area than the 1-bit configuration, improves efficiency to 1:24x over the
bit-parallel baseline while being 73% faster for convolutional layers and 60%
faster for fully-connected layers is also presented.
| arxiv topic:cs.NE |
arxiv_dataset-87421707.09168 | Learning to Predict Charges for Criminal Cases with Legal Basis
cs.CL
The charge prediction task is to determine appropriate charges for a given
case, which is helpful for legal assistant systems where the user input is fact
description. We argue that relevant law articles play an important role in this
task, and therefore propose an attention-based neural network method to jointly
model the charge prediction task and the relevant article extraction task in a
unified framework. The experimental results show that, besides providing legal
basis, the relevant articles can also clearly improve the charge prediction
results, and our full model can effectively predict appropriate charges for
cases with different expression styles.
| arxiv topic:cs.CL |
arxiv_dataset-87431707.09268 | Non-equilibrium evolution of volatility in origination and extinction
explains fat-tailed fluctuations in Phanerozoic biodiversity
q-bio.PE
Fluctuations in biodiversity, large and small, are pervasive in the fossil
record, yet we do not understand the processes generating them. Here we extend
theory from non-equilibrium statistical physics to describe the previously
unaccounted for fat-tailed form of fluctuations in marine invertebrate richness
through the Phanerozoic. Using this theory, known as superstatistics, we show
that the simple fact of heterogeneous origination and extinction rates between
clades and conserved rates within clades is sufficient to account for this
fat-tailed form. We identify orders and the families they subsume as the
taxonomic level at which clades experience inter-clade heterogeneity and within
clade homogeneity of rates. Following superstatistics we would posit that
orders and families are subsystems in local statistical equilibrium while the
entire system is not in equilibrium. The separation of timescales between
background origination and extinction within clades compared to the origin of
major ecological and evolutionary innovations leading to new clades allows
within-clade dynamics to reach equilibrium, while between-clade diversification
is non-equilibrial. This between clade non-equilibrium accounts for the
fat-tailed nature of the system as a whole. The distribution of shifts in
diversification dynamics across orders and families is consistent with niche
conservatism and pulsed exploration of adaptive landscapes by higher taxa.
Compared to other approaches that have used simple birth-death processes,
simple equilibrial dynamics, or non-linear theories from complexity science,
superstatistics is superior in its ability to account for both small and
extreme fluctuations in the richness of fossil taxa. Its success opens up new
research directions to better understand the evolutionary processes leading to
stasis in an adaptive landscape interrupted by innovations that lead to novel
forms.
| arxiv topic:q-bio.PE |
arxiv_dataset-87441707.09368 | Comments on my papers
math.RT
This document contains a description of several of my papers, including
remarks on history and connection with subsequent work. It also contains some
new results and conjectures.
| arxiv topic:math.RT |
arxiv_dataset-87451707.09468 | Zero-Shot Activity Recognition with Verb Attribute Induction
cs.CL cs.CV
In this paper, we investigate large-scale zero-shot activity recognition by
modeling the visual and linguistic attributes of action verbs. For example, the
verb "salute" has several properties, such as being a light movement, a social
act, and short in duration. We use these attributes as the internal mapping
between visual and textual representations to reason about a previously unseen
action. In contrast to much prior work that assumes access to gold standard
attributes for zero-shot classes and focuses primarily on object attributes,
our model uniquely learns to infer action attributes from dictionary
definitions and distributed word representations. Experimental results confirm
that action attributes inferred from language can provide a predictive signal
for zero-shot prediction of previously unseen activities.
| arxiv topic:cs.CL cs.CV |
arxiv_dataset-87461707.09568 | Collision of impurities with Bose-Einstein condensates
cond-mat.quant-gas
Quantum dynamics of impurities in a bath of bosons is a long-standing problem
of solid-state, plasma, and atomic physics. Recent experimental and theoretical
investigations with ultracold atoms focused on this problem, studying atomic
impurities immersed in a atomic Bose-Einstein condensate (BEC) and for various
relative coupling strengths tuned by the Fano-Feshbach resonance technique.
Here we report extensive numerical simulations on a closely related problem:
the collision between a bosonic impurity made of few $^{41}$K atoms and a BEC
made of $^{87}$Rb atoms in a quasi one-dimensional configuration and under a
weak harmonic axial confinement. For small values of the interspecies
interaction strength (no matter the sign of it), we find that the impurity,
which starts from outside the BEC, simply oscillates back and forth the BEC
cloud, but the frequency of oscillation depends on the interaction strength.
For intermediate couplings, after a few cycles of oscillation the impurity is
captured by the BEC and strongly changes its amplitude of oscillation. In the
strong interaction regime, if the interspecies interaction is attractive, a
local maximum (bright soliton) in the density of BEC occurs where the impurity
is trapped; instead, if the interspecies interaction is repulsive, the impurity
is not able to enter in the BEC cloud and the reflection coefficient is close
to one. On the other hand, if the initial displacement of the impurity is
increased, the impurity is able to penetrate in the cloud leading to the
appearance of a moving hole (dark soliton) in the BEC.
| arxiv topic:cond-mat.quant-gas |
arxiv_dataset-87471707.09668 | Handling Nested Parallelism and Extreme Load Imbalance in an Orbital
Analysis Code
cs.DC
Nested parallelism exists in scientific codes that are searching
multi-dimensional spaces. However, implementations of nested parallelism often
have overhead and load balance issues. The Orbital Analysis code we present
exhibits a sparse search space, significant load imbalances, and stopping when
the first solution is reached. All these aspects of the algorithm exacerbate
the problem of using nested parallelism effectively. In this paper, we present
an inspector/executor strategy for chunking such computations into parallel
wavefronts. The presented shared memory parallelization is no longer nested and
exhibits significantly less load imbalance. We evaluate this approach on an
Orbital analysis code, and we improve the execution time from the original
implementation by an order of magnitude. As part of a Graduate Computer Science
course in Parallel Programming models, we show how the approach can be
implemented in parallel Perl, Python, Chapel, Pthreads, and OpenMP. Future work
includes investigating how to automate and generalize the parallelization
approach.
| arxiv topic:cs.DC |
arxiv_dataset-87481707.09768 | Spin wave beam propagation in ferromagnetic thin film with graded
refractive index: mirage effect and prospective applications
cond-mat.mes-hall
Using analysis of iso-frequency contours of the spin-wave dispersion
relation, supported by micromagnetic simulations, we study the propagation of
spin-wave (SW) beams in thin ferromagnetic films through the areas of the
inhomogeneous refractive index. We compare the transmission and reflection of
SWs in areas with gradual and step variation of the SW refractive index. In
particular, we show the mirage effect for SWs with narrowing SW beam width, and
an application of the gradual modulation of the SWs refractive index as a
diverging lens. Furthermore, we study the propagation of SWs in ferromagnetic
stripe with modulated refractive index. We demonstrate that the system can be
considered as the graded-index waveguide, which preserves the width of the SW
beam for a long distance-the property essential for prospective applications of
magnonics.
| arxiv topic:cond-mat.mes-hall |
arxiv_dataset-87491707.09868 | Magnetism from intermetallics and perovskite oxides
cond-mat.mtrl-sci cond-mat.other
This work has been presented by RJCV to obtain his PhD degree at Fluminense
Federal University, in March of 2017. We focused on the synthesis of compounds
and then on their magneto-strucutral characterization; mainly due to the
interplay of these physical properties. We have prepared intermetallic alloys
(including Heusler alloys) and perovskite oxides (manganites and cobaltites);
in bulk and nanoparticles. A thorough analysis of the influence of the
morphology and crystal structure on the magnetic properties of these compounds
is addressed.
| arxiv topic:cond-mat.mtrl-sci cond-mat.other |
arxiv_dataset-87501707.09968 | Bounds on the burning numbers of spiders and path-forests
math.CO
Graph burning is one model for the spread of memes and contagion in social
networks. The corresponding graph parameter is the burning number of a graph
$G$, written $b(G)$, which measures the speed of the social contagion. While it
is conjectured that the burning number of a connected graph of order $n$ is at
most $\lceil \sqrt{n} \rceil$, this remains open in general and in many graph
families. We prove the conjectured bound for spider graphs, which are trees
with exactly one vertex of degree at least 3. To prove our result for spiders,
we develop new bounds on the burning number for path-forests, which in turn
leads to a $\frac 3 2$-approximation algorithm for computing the burning number
of path-forests.
| arxiv topic:math.CO |
arxiv_dataset-87511708.00088 | Learning Algorithms for Active Learning
cs.LG
We introduce a model that learns active learning algorithms via metalearning.
For a distribution of related tasks, our model jointly learns: a data
representation, an item selection heuristic, and a method for constructing
prediction functions from labeled training sets. Our model uses the item
selection heuristic to gather labeled training sets from which to construct
prediction functions. Using the Omniglot and MovieLens datasets, we test our
model in synthetic and practical settings.
| arxiv topic:cs.LG |
arxiv_dataset-87521708.00188 | On outer-connected domination for graph products
cs.DM
An outer-connected dominating set for an arbitrary graph $G$ is a set
$\tilde{D} \subseteq V$ such that $\tilde{D}$ is a dominating set and the
induced subgraph $G [V \setminus \tilde{D}]$ be connected. In this paper, we
focus on the outer-connected domination number of the product of graphs. We
investigate the existence of outer-connected dominating set in lexicographic
product and Corona of two arbitrary graphs, and we present upper bounds for
outer-connected domination number in lexicographic and Cartesian product of
graphs. Also, we establish an equivalent form of the Vizing's conjecture for
outer-connected domination number in lexicographic and Cartesian product as
$\tilde{\gamma_c}(G \circ K)\tilde{\gamma_c}(H \circ K) \leq
\tilde{\gamma_c}(G\Box H)\circ K$. Furthermore, we study the outer-connected
domination number of the direct product of finitely many complete graphs.
| arxiv topic:cs.DM |
arxiv_dataset-87531708.00288 | Rare-earth/transition-metal magnetic interactions in pristine and
(Ni,Fe)-doped YCo5 and GdCo5
cond-mat.mtrl-sci
We present an investigation into the intrinsic magnetic properties of the
compounds YCo5 and GdCo5, members of the RETM5 class of permanent magnets (RE =
rare earth, TM = transition metal). Focusing on Y and Gd provides direct
insight into both the TM magnetization and RE-TM interactions without the
complication of strong crystal field effects. We synthesize single crystals of
YCo5 and GdCo5 using the optical floating zone technique and measure the
magnetization from liquid helium temperatures up to 800 K. These measurements
are interpreted through calculations based on a Green's function formulation of
density-functional theory, treating the thermal disorder of the local magnetic
moments within the coherent potential approximation. The rise in the
magnetization of GdCo5 with temperature is shown to arise from a faster
disordering of the Gd magnetic moments compared to the antiferromagnetically
aligned Co sublattice. We use the calculations to analyze the different Curie
temperatures of the compounds and also compare the molecular (Weiss) fields at
the RE site with previously published neutron scattering experiments. To gain
further insight into the RE-TM interactions, we perform substitutional doping
on the TM site, studying the compounds RECo4.5Ni0.5, RECo4Ni, and RECo4.5Fe0.5.
Both our calculations and experiments on powdered samples find an
increased/decreased magnetization with Fe/Ni doping, respectively. The
calculations further reveal a pronounced dependence on the location of the
dopant atoms of both the Curie temperatures and the Weiss field at the RE site.
| arxiv topic:cond-mat.mtrl-sci |
arxiv_dataset-87541708.00388 | The Vela X pulsar wind nebula through the eyes of H.E.S.S. and Suzaku
astro-ph.HE
Pulsar wind nebulae (PWNe) are among the most extreme particle accelerators
in galaxies, are recognized as multi-TeV electron/positron sources, and are one
of the dominant classes of Galactic gamma-ray sources. Vela X is a nearby PWN
at 290 pc from the Earth with large apparent size ($>1^\circ$). The H.E.S.S.
array of imaging atmospheric Cherenkov telescopes has detected Vela X as one of
the brightest known sources of TeV gamma rays. The bulk of the gamma-ray
emission measured using H.E.S.S. coincides with an elongated structure known
from X-ray observations and dubbed the cocoon, that seemingly emanates from the
region of the pulsar wind termination shock. The spectral energy distribution
of the cocoon peaks at around 10 TeV, and then presents a cutoff that can be
precisely measured with H.E.S.S. owing to the extreme brightness of the source.
Electrons radiating inverse-Compton gamma rays in the cutoff region are the
same responsible for the X-ray synchrotron emission at energies $> 1$ keV.
Therefore, Vela X provides a unique test case, in which we can constrain the
densities and spectra of accelerated leptons in the cutoff regime, as well as
the magnetic field properties, with minimal modeling assumptions. Thanks to the
proximity/large apparent size of the source, this can be done in a
spatially-resolved fashion across the PWN. We will present an analysis of
H.E.S.S. data combined with X-ray data from the Suzaku space telescope. We will
discuss implications for the mechanisms behind particle acceleration and
transport, constrain the strength of the magnetic field in different locations
in the nebula, and probe for magnetic field turbulence.
| arxiv topic:astro-ph.HE |
arxiv_dataset-87551708.00488 | A Second Order Ensemble Timestepping Algorithm for Natural Convection
math.NA
This paper presents an algorithm for calculating an ensemble of solutions to
natural convection problems. The ensemble average is the most likely
temperature distribution and its variance gives an estimate of prediction
reliability. Solutions are calculated by solving two coupled linear systems,
each involving a shared coefficient matrix, for multiple right-hand sides at
each timestep. Storage requirements and computational costs to solve the system
are thereby reduced. Moreover, this paper addresses a need for higher order
methods to solve natural convection problems. Stability and convergence of the
method are proven under a timestep condition involving fluctuations of the
velocity. Numerical tests are provided which confirm the theoretical analyses.
| arxiv topic:math.NA |
arxiv_dataset-87561708.00588 | Hidden Physics Models: Machine Learning of Nonlinear Partial
Differential Equations
cs.AI cs.LG math.AP stat.ML
While there is currently a lot of enthusiasm about "big data", useful data is
usually "small" and expensive to acquire. In this paper, we present a new
paradigm of learning partial differential equations from {\em small} data. In
particular, we introduce \emph{hidden physics models}, which are essentially
data-efficient learning machines capable of leveraging the underlying laws of
physics, expressed by time dependent and nonlinear partial differential
equations, to extract patterns from high-dimensional data generated from
experiments. The proposed methodology may be applied to the problem of
learning, system identification, or data-driven discovery of partial
differential equations. Our framework relies on Gaussian processes, a powerful
tool for probabilistic inference over functions, that enables us to strike a
balance between model complexity and data fitting. The effectiveness of the
proposed approach is demonstrated through a variety of canonical problems,
spanning a number of scientific domains, including the Navier-Stokes,
Schr\"odinger, Kuramoto-Sivashinsky, and time dependent linear fractional
equations. The methodology provides a promising new direction for harnessing
the long-standing developments of classical methods in applied mathematics and
mathematical physics to design learning machines with the ability to operate in
complex domains without requiring large quantities of data.
| arxiv topic:cs.AI cs.LG math.AP stat.ML |
arxiv_dataset-87571708.00688 | Contact angles of liquid drops subjected to a rough boundary
math-ph math.MP
The contact angle of a liquid drop on a rigid surface is determined by the
classical theory of Young-Laplace. For chemically homogeneous surfaces, this
angle is a constant.
We study the minimal-energy configurations of liquid drops on rough surfaces.
Here the actual angle is still constant for homogeneous surfaces, but the
apparent angle can fluctuate widely. A limit theorem is introduced for minimal
energy configuration, where the rigid surface converges to a smooth one, but
the roughness parameter is kept constant. It turns out that the limit of
minimal energy configurations correspond to liquid drop on a smooth surface
with an appropriately defined effective chemical interaction energy. It turns
out that the effective chemical interaction depends linearly on the roughness
in a certain range of parameters, corresponding to full wetting. Outside this
range the most stable configuration corresponds to a partial wetting and the
effective interaction energy depends on the geometry in an essential way.
This result partially justifies and extends Wenzel and Cassie's laws and can
be used to deduce the actual inclination angle in the most stable state, where
the apparent one is known by measurement. This, in turn, may be applied to
deduce the roughness parameter if the interfacial energy is known, or visa
versa.
| arxiv topic:math-ph math.MP |
arxiv_dataset-87581708.00788 | A generalized Schwarz lemma for two domains related to $\mu$-synthesis
math.CV
We present a set of necessary and sufficient conditions that provides a
Schwarz lemma for the tetrablock $\mathbb E$. As an application of this result,
we obtain a Schwarz lemma for the symmetrized bidisc $\mathbb G_2$. In either
case, our results generalize all previous results in this direction for
$\mathbb E$ and $\mathbb G_2$.
| arxiv topic:math.CV |
arxiv_dataset-87591708.00888 | Partonic Structure of Light Nuclei
nucl-ex
We propose to study the partonic structure of $^4$He by measuring the Beam
Spin Asymmetry (BSA) in coherent Deeply Virtual Compton Scattering (DVCS) and
the differential cross-section of the Deeply Virtual Meson Production (DVMP) of
the $\phi$. Despite its simple structure, a light nucleus such as $^4$He has a
density and a binding energy comparable to that of heavier nuclei. Therefore,
by studying $^4$He nucleus, one can learn typical features of the partonic
structure of atomic nuclei.
The combination of CLAS12 and the ALERT detector provides a unique
opportunity to study both the quark and gluon structure of a dense light
nucleus. Coherent exclusive DVCS off $^4$He will probe the transverse spatial
distribution of quarks in the nucleus as a function of the quarks' longitudinal
momentum fraction, $x$. In parallel, the average spatial transverse gluon
density of the $^4$He nucleus will be extracted within a GPD framework using
the measured longitudinal cross-section for coherent $\phi$ production in a
similar range of $x$. Additionally, threshold effects of $\phi$ production can
be explored by exploiting the ALERT detector's large acceptance for low $|t|$
events.
| arxiv topic:nucl-ex |
arxiv_dataset-87601708.00988 | A study of the morphology, dynamics, and folding pathways of ring
polymers with supramolecular topological constraints using molecular
simulation and nonlinear manifold learning
physics.chem-ph physics.data-an
Ring polymers are prevalent in natural and engineered systems, including
circular bacterial DNA, crown ethers for cation chelation, and mechanical
nanoswitches. The morphology and dynamics of ring polymers are governed by the
chemistry and degree of polymerization of the ring, and intramolecular and
supramolecular topological constraints such as knots or
mechanically-interlocked rings. In this study, we perform molecular dynamics
simulations of polyethylene ring polymers at two different degrees of
polymerization and in different topological states, including a trefoil knot,
catenane state (two interlocked rings), and Borromean state (three interlocked
rings). We employ nonlinear manifold learning to extract the low-dimensional
free energy surface to which the structure and dynamics of the polymer chain
are effectively restrained. The free energy surfaces reveal how degree of
polymerization and topological constraints affect the thermally accessible
conformations, chiral symmetry breaking, and folding and collapse pathways of
the rings, and present a means to rationally engineer ring size and topology to
stabilize particular conformational states and folding pathways. We compute the
rotational diffusion of the ring in these various states as a crucial property
required for the design of engineered devices containing ring polymer
components.
| arxiv topic:physics.chem-ph physics.data-an |
arxiv_dataset-87611708.01088 | GRB Observations with H.E.S.S. II
astro-ph.HE
The High Energy Stereoscopic System (H.E.S.S.) has been searching for
counterparts of Gamma Ray Bursts (GRBs) for many years. In 2012 the system was
upgraded with a fifth $28$ m diameter telescope (CT5) which is equipped with
faster motors for rapid repointing, marking the start of the second phase of
H.E.S.S. operation (H.E.S.S. II). CT5s large light collection area of
$600\,{\rm m}^{2}$ improves the sensitivity to low-energy gamma-rays and even
extends the energy range below $100$ GeV. The search for counterparts continues
now in the energy range of tens of GeV to tens of TeV. A detection in this
energy range would open a new window to the part of the spectrum of these
highly energetic explosions which Fermi-LAT has only successfully detected in a
reduced subset of events, with rather limited statistics. In the past years,
H.E.S.S. has performed followup observations based on GRB detections by
Swift-BAT and Fermi-GBM/-LAT. This Target of Opportunity observation program
was carried out with a generalised Target of Opportunity Alert system. This
contribution will highlight key features of the Target of Opportunity Alert
system, present follow-up statistics of GRBs as well as detailed results of
promising follow-up observations.
| arxiv topic:astro-ph.HE |
arxiv_dataset-87621708.01188 | Hardy Spaces ($1<p<\infty$) over Lipschitz Domains
math.CV
Let $\Gamma$ be a Lipschitz curve on the complex plane $\mathbb{C}$ and
$\Omega_+$ is the domain above $\Gamma$, we define Hardy space $H^p(\Omega_+)$
as the set of holomorphic functions $F$ satisfying $\sup_{\tau>0}(\int_{\Gamma}
|F(\zeta+\mathrm{i}\tau)|^p |\,\mathrm{d}\zeta|)^{\frac1p}< \infty$. We mainly
focus on the case of $1<p<\infty$ in this paper, and prove that if $F(w)\in
H^p(\Omega_+)$, then $F(w)$ has non-tangential boundary limit $F(\zeta)$ a.e.
on $\Gamma$, and $F(w)$ is the Cauchy integral of $F(\zeta)$. We denote the
conformal mapping from $\mathbb{C}_+$ onto $\Omega_+$ as $\Phi$, and then prove
that, $ H^p(\Omega_+)$ is isomorphic to $H^p(\mathbb{C}_+)$, the classical
Hardy space on upper half plane, under the mapping $T\colon F\to
F(\Phi(z))\cdot (\Phi'(z))^\frac{1}{p}$, where $F\in H^p(\Omega_+)$.
| arxiv topic:math.CV |
arxiv_dataset-87631708.01288 | Twist star products and Morita equivalence
math.QA math-ph math.MP
We present a simple no-go theorem for the existence of a deformation
quantization of a homogeneous space M induced by a Drinfel'd twist: we argue
that equivariant line bundles on M with non-trivial Chern class and symplectic
twist star products cannot both exist on the same manifold M. This implies, for
example, that there is no symplectic star product on the complex projective
spaces induced by a twist based on U(gl(n,C))[[h]] or any sub-bialgebra, for
every n greater or equal than 2.
| arxiv topic:math.QA math-ph math.MP |
arxiv_dataset-87641708.01388 | Performance Overhead Comparison between Hypervisor and Container based
Virtualization
cs.DC
The current virtualization solution in the Cloud widely relies on
hypervisor-based technologies. Along with the recent popularity of Docker, the
container-based virtualization starts receiving more attention for being a
promising alternative. Since both of the virtualization solutions are not
resource-free, their performance overheads would lead to negative impacts on
the quality of Cloud services. To help fundamentally understand the performance
difference between these two types of virtualization solutions, we use a
physical machine with "just-enough" resource as a baseline to investigate the
performance overhead of a standalone Docker container against a standalone
virtual machine (VM). With findings contrary to the related work, our
evaluation results show that the virtualization's performance overhead could
vary not only on a feature-by-feature basis but also on a job-to-job basis.
Although the container-based solution is undoubtedly lightweight, the
hypervisor-based technology does not come with higher performance overhead in
every case. For example, Docker containers particularly exhibit lower QoS in
terms of storage transaction speed.
| arxiv topic:cs.DC |
arxiv_dataset-87651708.01488 | Focusing RKKY interaction by graphene P-N junction
cond-mat.mes-hall
The carrier-mediated RKKY interaction between local spins plays an important
role for the application of magnetically doped graphene in spintronics and
quantum computation. Previous studies largely concentrate on the influence of
electronic states of uniform systems on the RKKY interaction. Here we reveal a
very different way to manipulate the RKKY interaction by showing that the
anomalous focusing - a well-known electron optics phenomenon in graphene P-N
junctions - can be utilized to refocus the massless Dirac electrons emanating
from one local spin to the other local spin. This gives rise to rich spatial
interference patterns and symmetry-protected non-oscillatory RKKY interaction
with a strongly enhanced magnitude. It may provide a new way to engineer the
long-range spin-spin interaction in graphene.
| arxiv topic:cond-mat.mes-hall |
arxiv_dataset-87661708.01588 | Fredholm Theory and Optimal Test Functions for Detecting Central Point
Vanishing Over Families of L-functions
math.NT
The Riemann Zeta-Function is the most studied L-function; it's zeroes give
information about the prime numbers. We can associate L-functions to a wide
array of objects, and in general, the zeroes of these L-functions give
information about those objects. For arbitrary L-functions, the order of
vanishing at the central point is of particular important. For example, the
Birch and Swinnerton-Dyer conjecture states that the order of vanishing at the
central point of an elliptic curve L-function is the rank of the Mordell-Weil
group of that elliptic curve.
The Katz-Sarnak Density Conjecture states that this order vanishing and other
behavior are well-modeled by random matrices drawn from the classical compact
groups. In particular, the conjecture states that an average order vanishing
over a family of L-functions can be bounded using only a given weight function
and a chosen test function, phi. The conjecture is known for many families when
the test functions are suitably restricted.
It is natural to ask which test function is best for each family and for each
set of natural restrictions on phi. Our main result is a reduction of an
otherwise infinite dimensional optimization to a finite-dimensional
optimization problem for all families and all sets of restrictions. We
explicitly solve many of these optimization problems and compute the improved
bound we obtain on average rank. While we do not verify the density conjecture
for these new, looser restrictions, with this project, we are able to precisely
quantify the benefits of such efforts with respect to average rank. Finally, we
are able to show that this bound strictly improves as we increase support.
| arxiv topic:math.NT |
arxiv_dataset-87671708.01688 | Abstract Hidden Markov Models: a monadic account of quantitative
information flow
cs.LO
Hidden Markov Models, HMM's, are mathematical models of Markov processes with
state that is hidden, but from which information can leak. They are typically
represented as 3-way joint-probability distributions.
We use HMM's as denotations of probabilistic hidden-state sequential
programs: for that, we recast them as `abstract' HMM's, computations in the
Giry monad $\mathbb{D}$, and we equip them with a partial order of increasing
security. However to encode the monadic type with hiding over some state
$\mathcal{X}$ we use $\mathbb{D}\mathcal{X}\to \mathbb{D}^2\mathcal{X}$ rather
than the conventional $\mathcal{X}{\to}\mathbb{D}\mathcal{X}$ that suffices for
Markov models whose state is not hidden. We illustrate the
$\mathbb{D}\mathcal{X}\to \mathbb{D}^2\mathcal{X}$ construction with a small
Haskell prototype.
We then present uncertainty measures as a generalisation of the extant
diversity of probabilistic entropies, with characteristic analytic properties
for them, and show how the new entropies interact with the order of increasing
security. Furthermore, we give a `backwards' uncertainty-transformer semantics
for HMM's that is dual to the `forwards' abstract HMM's - it is an analogue of
the duality between forwards, relational semantics and backwards,
predicate-transformer semantics for imperative programs with demonic choice.
Finally, we argue that, from this new denotational-semantic viewpoint, one
can see that the Dalenius desideratum for statistical databases is actually an
issue in compositionality. We propose a means for taking it into account.
| arxiv topic:cs.LO |
arxiv_dataset-87681708.01788 | Neutrino scattering in supernovae and spin correlations of a unitary gas
astro-ph.HE cond-mat.quant-gas nucl-th
Core collapse supernova simulations can be sensitive to neutrino interactions
near the neutrinosphere. This is the surface of last scattering. We model the
neutrinosphere region as a warm unitary gas of neutrons. A unitary gas is a low
density system of particles with large scattering lengths. We calculate
modifications to neutrino scattering cross sections because of the universal
spin and density correlations of a unitary gas. These correlations can be
studied in laboratory cold atom experiments. We find significant reductions in
cross sections, compared to free space interactions, even at relatively low
densities. These reductions could reduce the delay time from core bounce to
successful explosion in multidimensional supernova simulations.
| arxiv topic:astro-ph.HE cond-mat.quant-gas nucl-th |
arxiv_dataset-87691708.01888 | Connectivity Inference from Neural Recording Data: Challenges,
Mathematical Bases and Research Directions
q-bio.NC q-bio.QM
This article presents a review of computational methods for connectivity
inference from neural activity data derived from multi-electrode recordings or
fluorescence imaging. We first identify biophysical and technical challenges in
connectivity inference along the data processing pipeline. We then review
connectivity inference methods based on two major mathematical foundations,
namely, descriptive model-free approaches and generative model-based
approaches. We investigate representative studies in both categories and
clarify which challenges have been addressed by which method. We further
identify critical open issues and possible research directions.
| arxiv topic:q-bio.NC q-bio.QM |
arxiv_dataset-87701708.01988 | Identity-Aware Textual-Visual Matching with Latent Co-attention
cs.CV
Textual-visual matching aims at measuring similarities between sentence
descriptions and images. Most existing methods tackle this problem without
effectively utilizing identity-level annotations. In this paper, we propose an
identity-aware two-stage framework for the textual-visual matching problem. Our
stage-1 CNN-LSTM network learns to embed cross-modal features with a novel
Cross-Modal Cross-Entropy (CMCE) loss. The stage-1 network is able to
efficiently screen easy incorrect matchings and also provide initial training
point for the stage-2 training. The stage-2 CNN-LSTM network refines the
matching results with a latent co-attention mechanism. The spatial attention
relates each word with corresponding image regions while the latent semantic
attention aligns different sentence structures to make the matching results
more robust to sentence structure variations. Extensive experiments on three
datasets with identity-level annotations show that our framework outperforms
state-of-the-art approaches by large margins.
| arxiv topic:cs.CV |
arxiv_dataset-87711708.02088 | Finite subgraphs of an extension graph
math.GR math.GT
Let $\Gamma$ be a finite graph and let $\Gamma^{\mathrm{e}}$ be its extension
graph. We inductively define a sequence $\{\Gamma_i\}$ of finite induced
subgraphs of $\Gamma^{\mathrm{e}}$ through successive applications of an
operation called "doubling along a star". Then we show that every finite
induced subgraph of $\Gamma^{\mathrm{e}}$ is isomorphic to an induced subgraph
of some $\Gamma_i$.
| arxiv topic:math.GR math.GT |
arxiv_dataset-87721708.02188 | PowerAI DDL
cs.DC cs.AI cs.LG
As deep neural networks become more complex and input datasets grow larger,
it can take days or even weeks to train a deep neural network to the desired
accuracy. Therefore, distributed Deep Learning at a massive scale is a critical
capability, since it offers the potential to reduce the training time from
weeks to hours. In this paper, we present a software-hardware co-optimized
distributed Deep Learning system that can achieve near-linear scaling up to
hundreds of GPUs. The core algorithm is a multi-ring communication pattern that
provides a good tradeoff between latency and bandwidth and adapts to a variety
of system configurations. The communication algorithm is implemented as a
library for easy use. This library has been integrated into Tensorflow, Caffe,
and Torch. We train Resnet-101 on Imagenet 22K with 64 IBM Power8 S822LC
servers (256 GPUs) in about 7 hours to an accuracy of 33.8 % validation
accuracy. Microsoft's ADAM and Google's DistBelief results did not reach 30 %
validation accuracy for Imagenet 22K. Compared to Facebook AI Research's recent
paper on 256 GPU training, we use a different communication algorithm, and our
combined software and hardware system offers better communication overhead for
Resnet-50. A PowerAI DDL enabled version of Torch completed 90 epochs of
training on Resnet 50 for 1K classes in 50 minutes using 64 IBM Power8 S822LC
servers (256 GPUs).
| arxiv topic:cs.DC cs.AI cs.LG |
arxiv_dataset-87731708.02288 | Beyond Low-Rank Representations: Orthogonal Clustering Basis
Reconstruction with Optimized Graph Structure for Multi-view Spectral
Clustering
cs.CV
Low-Rank Representation (LRR) is arguably one of the most powerful paradigms
for Multi-view spectral clustering, which elegantly encodes the multi-view
local graph/manifold structures into an intrinsic low-rank self-expressive data
similarity embedded in high-dimensional space, to yield a better graph
partition than their single-view counterparts. In this paper we revisit it with
a fundamentally different perspective by discovering LRR as essentially a
latent clustered orthogonal projection based representation winged with an
optimized local graph structure for spectral clustering; each column of the
representation is fundamentally a cluster basis orthogonal to others to
indicate its members, which intuitively projects the view-specific feature
representation to be the one spanned by all orthogonal basis to characterize
the cluster structures. Upon this finding, we propose our technique with the
followings: (1) We decompose LRR into latent clustered orthogonal
representation via low-rank matrix factorization, to encode the more flexible
cluster structures than LRR over primal data objects; (2) We convert the
problem of LRR into that of simultaneously learning orthogonal clustered
representation and optimized local graph structure for each view; (3) The
learned orthogonal clustered representations and local graph structures enjoy
the same magnitude for multi-view, so that the ideal multi-view consensus can
be readily achieved. The experiments over multi-view datasets validate its
superiority.
| arxiv topic:cs.CV |
arxiv_dataset-87741708.02388 | Multipole excitations in hot nuclei within the finite temperature
quasiparticle random phase approximation framework
nucl-th
The effect of temperature on the evolution of the isovector dipole and
isoscalar quadrupole excitations in $^{68}$Ni and $^{120}$Sn nuclei is studied
within the fully self-consistent finite temperature quasiparticle random phase
approximation framework, based on the Skyrme-type SLy5 energy density
functional. The new low-energy excitations emerge due to the transitions from
thermally occupied states to the discretized continuum at finite temperatures,
whereas the isovector giant dipole resonance is not strongly impacted by the
increase of temperature. The radiative dipole strength at low-energies is also
investigated for the $^{122}$Sn nucleus, becoming compatible with the available
experimental data when the temperature is included. In addition, both the
isoscalar giant quadrupole resonance and low-energy quadrupole states are
sensitive to the temperature effect: while the centroid energies decrease in
the case of the isoscalar giant quadrupole resonance, the collectivity of the
first $2^{+}$ state is quenched and the opening of new excitation channels
fragments the low-energy strength at finite temperatures.
| arxiv topic:nucl-th |
arxiv_dataset-87751708.02488 | Convergence analysis of Riemannian Gauss-Newton methods and its
connection with the geometric condition number
math.NA
We obtain estimates of the multiplicative constants appearing in local
convergence results of the Riemannian Gauss-Newton method for least squares
problems on manifolds and relate them to the geometric condition number of [P.
B\"urgisser and F. Cucker, Condition: The Geometry of Numerical Algorithms,
2013].
| arxiv topic:math.NA |
arxiv_dataset-87761708.02588 | Effects of primordial black holes quantum gravity decay on galaxy
clustering
astro-ph.CO gr-qc
It has been recently suggested that small mass black holes (BHs) may become
unstable due to quantum-gravitational effects and eventually decay, producing
radiation, on a timescale shorter than the Hawking evaporation time. We argue
that the existence of a population of low-mass Primordial Black Holes (PBHs)
acting as a fraction of the Universe dark matter component can be used to test
proposed models of quantum decay of BHs via their effect on galaxy number
counts. We study what constraints future galaxy clustering measurements can set
on quantum-gravity parameters governing the BH lifetime and PBH abundance. In
case of no detection of such effects, this would rule out either the existence
of a non-negligible number of small PBHs, or the BH quantum decay scenario (or
both). In case of independent observations of PBHs, the observables discussed
here could be used to study the quantum effects that modify the final fate of
BHs.
| arxiv topic:astro-ph.CO gr-qc |
arxiv_dataset-87771708.02688 | Statistics of Deep Generated Images
cs.CV
Here, we explore the low-level statistics of images generated by
state-of-the-art deep generative models. First, Variational auto-encoder
(VAE~\cite{kingma2013auto}), Wasserstein generative adversarial network
(WGAN~\cite{arjovsky2017wasserstein}) and deep convolutional generative
adversarial network (DCGAN~\cite{radford2015unsupervised}) are trained on the
ImageNet dataset and a large set of cartoon frames from animations. Then, for
images generated by these models as well as natural scenes and cartoons,
statistics including mean power spectrum, the number of connected components in
a given image area, distribution of random filter responses, and contrast
distribution are computed. Our analyses on training images support current
findings on scale invariance, non-Gaussianity, and Weibull contrast
distribution of natural scenes. We find that although similar results hold over
cartoon images, there is still a significant difference between statistics of
natural scenes and images generated by VAE, DCGAN and WGAN models. In
particular, generated images do not have scale invariant mean power spectrum
magnitude, which indicates existence of extra structures in these images.
Inspecting how well the statistics of deep generated images match the known
statistical properties of natural images, such as scale invariance,
non-Gaussianity, and Weibull contrast distribution, can a) reveal the degree to
which deep learning models capture the essence of the natural scenes, b)
provide a new dimension to evaluate models, and c) allow possible improvement
of image generative models (e.g., via defining new loss functions).
| arxiv topic:cs.CV |
arxiv_dataset-87781708.02788 | Fractional powers of the parabolic Hermite operator. Regularity
properties
math.FA
Let $\mathcal{L}= \partial_t- \Delta_x+|x|^2$. Consider its Poisson semigroup
$e^{-y\sqrt{\mathcal{L}}}$. For $\alpha >0$ define the Parabolic
Hermite-Zygmund spaces
$$
\Lambda^\alpha_{\mathcal{L}}=\left\{f: \:f\in L^\infty(\mathbb{R}^{n+1})\:\;
{\rm and} \:\; \left\|\partial_y^k e^{-y\sqrt{\mathcal{L}}} f
\right\|_{L^\infty(\mathbb{R}^{n+1})}\leq C_k y^{-k+\alpha},\;\: {\rm with }\,
k=[\alpha]+1, y>0. \right\},
$$
with the obvious norm. It is shown that these spaces have a pointwise
description of H\"older type.
The fractional powers $\mathcal{L}^{\pm \beta}$ are well defined in these
spaces and the following regularity properties are proved:
\begin{eqnarray*}
\alpha, \beta >0, \quad \|\mathcal{L}^{-\beta} f\|_{
\Lambda^{\alpha+2\beta}_{\mathcal{L}}}\le C \|f\|_{
\Lambda^\alpha_{\mathcal{L}}}.
\end{eqnarray*}
\begin{eqnarray*}
0< 2\beta < \alpha, \quad \|\mathcal{L}^\beta
f\|_{\Lambda_{\mathcal{L}}^{\alpha-2\beta}}\le C
\|f\|_{\Lambda^\alpha_{\mathcal{L}}}.
\end{eqnarray*}
Parallel results are obtained for the Hermite operator $- \Delta +|x|^2.$ The
proofs use in a fundamental way the semigroup definition of the operators
$\mathcal{L}^{\pm \beta}$ and $(-\Delta+|x|^2)^{\pm \beta}$. The
non-convolution structure of the operators produce an extra difficulty of the
arguments.
| arxiv topic:math.FA |
arxiv_dataset-87791708.02888 | Multi-message Authentication over Noisy Channel with Secure Channel
Codes
cs.CR cs.IT math.IT
In this paper, we investigate multi-message authentication to combat
adversaries with infinite computational capacity. An authentication framework
over a wiretap channel $(W_1,W_2)$ is proposed to achieve information-theoretic
security with the same key. The proposed framework bridges the two research
areas in physical (PHY) layer security: secure transmission and message
authentication. Specifically, the sender Alice first transmits message $M$ to
the receiver Bob over $(W_1,W_2)$ with an error correction code; then Alice
employs a hash function (i.e., $\varepsilon$-AWU$_2$ hash functions) to
generate a message tag $S$ of message $M$ using key $K$, and encodes $S$ to a
codeword $X^n$ by leveraging an existing strongly secure channel coding with
exponentially small (in code length $n$) average probability of error; finally,
Alice sends $X^n$ over $(W_1,W_2)$ to Bob who authenticates the received
messages. We develop a theorem regarding the requirements/conditions for the
authentication framework to be information-theoretic secure for authenticating
a polynomial number of messages in terms of $n$. Based on this theorem, we
propose an authentication protocol that can guarantee the security
requirements, and prove its authentication rate can approach infinity when $n$
goes to infinity. Furthermore, we design and implement an efficient and
feasible authentication protocol over binary symmetric wiretap channel (BSWC)
by using \emph{Linear Feedback Shifting Register} based (LFSR-based) hash
functions and strong secure polar code. Through extensive experiments, it is
demonstrated that the proposed protocol can achieve low time cost, high
authentication rate, and low authentication error rate.
| arxiv topic:cs.CR cs.IT math.IT |
arxiv_dataset-87801708.02988 | Wrinkled few-layer graphene as highly efficient load bearer
cond-mat.mtrl-sci
Multilayered graphitic materials are not suitable as load-bearers due to
their inherent weak interlayer bonding (for example, graphite is a solid
lubricant in certain applications). This situation is largely improved when
two-dimensional (2-D) materials such as a monolayer (SLG) graphene are
employed. The downside in these cases is the presence of thermally or
mechanically induced wrinkles which are ubiquitous in 2-D materials. Here we
set out to examine the effect of extensive large wavelength/ amplitude
wrinkling on the stress transfer capabilities of exfoliated simply-supported
graphene flakes. Contrary to common belief we present clear evidence that this
type of "corrugation" enhances the load bearing capacity of few-layer graphene
as compared to 'flat' specimens. This effect is the result of the significant
increase of the graphene/polymer interfacial shear stress per increment of
applied strain due to wrinkling and paves the way for designing affordable
graphene composites with highly improved stress-transfer efficiency.
| arxiv topic:cond-mat.mtrl-sci |
arxiv_dataset-87811708.03088 | Semantic Video CNNs through Representation Warping
cs.CV
In this work, we propose a technique to convert CNN models for semantic
segmentation of static images into CNNs for video data. We describe a warping
method that can be used to augment existing architectures with very little
extra computational cost. This module is called NetWarp and we demonstrate its
use for a range of network architectures. The main design principle is to use
optical flow of adjacent frames for warping internal network representations
across time. A key insight of this work is that fast optical flow methods can
be combined with many different CNN architectures for improved performance and
end-to-end training. Experiments validate that the proposed approach incurs
only little extra computational cost, while improving performance, when video
streams are available. We achieve new state-of-the-art results on the CamVid
and Cityscapes benchmark datasets and show consistent improvements over
different baseline networks. Our code and models will be available at
http://segmentation.is.tue.mpg.de
| arxiv topic:cs.CV |
arxiv_dataset-87821708.03188 | Shenfun -- automating the spectral Galerkin method
physics.comp-ph
With the shenfun Python module (github.com/spectralDNS/shenfun) an effort is
made towards automating the implementation of the spectral Galerkin method for
simple tensor product domains, consisting of (currently) one non-periodic and
any number of periodic directions. The user interface to shenfun is
intentionally made very similar to FEniCS (fenicsproject.org). Partial
Differential Equations are represented through weak variational forms and
solved using efficient direct solvers where available. MPI decomposition is
achieved through the {mpi4py-fft} module (bitbucket.org/mpi4py/mpi4py-fft), and
all developed solver may, with no additional effort, be run on supercomputers
using thousands of processors. Complete solvers are shown for the linear
Poisson and biharmonic problems, as well as the nonlinear and time-dependent
Ginzburg-Landau equation.
| arxiv topic:physics.comp-ph |
arxiv_dataset-87831708.03288 | Subset Selection with Shrinkage: Sparse Linear Modeling when the SNR is
low
stat.ME math.OC math.ST stat.CO stat.ML stat.TH
We study a seemingly unexpected and relatively less understood overfitting
aspect of a fundamental tool in sparse linear modeling - best subset selection,
which minimizes the residual sum of squares subject to a constraint on the
number of nonzero coefficients. While the best subset selection procedure is
often perceived as the "gold standard" in sparse learning when the signal to
noise ratio (SNR) is high, its predictive performance deteriorates when the SNR
is low. In particular, it is outperformed by continuous shrinkage methods, such
as ridge regression and the Lasso. We investigate the behavior of best subset
selection in the high-noise regimes and propose an alternative approach based
on a regularized version of the least-squares criterion. Our proposed
estimators (a) mitigate, to a large extent, the poor predictive performance of
best subset selection in the high-noise regimes; and (b) perform favorably,
while generally delivering substantially sparser models, relative to the best
predictive models available via ridge regression and the Lasso. We conduct an
extensive theoretical analysis of the predictive properties of the proposed
approach and provide justification for its superior predictive performance
relative to best subset selection when the noise-level is high. Our estimators
can be expressed as solutions to mixed integer second order conic optimization
problems and, hence, are amenable to modern computational tools from
mathematical optimization.
| arxiv topic:stat.ME math.OC math.ST stat.CO stat.ML stat.TH |
arxiv_dataset-87841708.03388 | Reproducing kernel functions and asymptotic expansions on Jordan-Kepler
manifolds
math.CV math-ph math.DG math.MP math.RT
We study the complex geometry of generalized Kepler manifolds, defined in
Jordan theoretic terms, introduce Hilbert spaces of holomorphic functions
defined by radial measures, and find the complete asymptotic expansion of the
corresponding reproducing kernels for K\"ahler potentials, both in the flat and
bounded setting.
| arxiv topic:math.CV math-ph math.DG math.MP math.RT |
arxiv_dataset-87851708.03488 | Dipole force free optical control and cooling of nanofiber trapped atoms
physics.atom-ph
The evanescent field surrounding nano-scale optical waveguides offers an
efficient interface between light and mesoscopic ensembles of neutral atoms.
However, the thermal motion of trapped atoms, combined with the strong radial
gradients of the guided light, leads to a time-modulated coupling between atoms
and the light mode, thus giving rise to additional noise and motional dephasing
of collective states. Here, we present a dipole force free scheme for coupling
of the radial motional states, utilizing the strong intensity gradient of the
guided mode and demonstrate all-optical coupling of the cesium hyperfine ground
states and motional sideband transitions. We utilize this to prolong the trap
lifetime of an atomic ensemble by Raman sideband cooling of the radial motion,
which has not been demonstrated in nano-optical structures previously. Our work
points towards full and independent control of internal and external atomic
degrees of freedom using guided light modes only.
| arxiv topic:physics.atom-ph |
arxiv_dataset-87861708.03588 | Canonical Field Anticommutators in the Extended Gauged Rarita-Schwinger
Theory
hep-th
We reexamine canonical quantization of the gauged Rarita-Schwinger theory
using the extended theory, incorporating a dimension $\frac{1}{2}$ auxiliary
spin-$\frac{1}{2}$ field $\Lambda$, in which there is an exact off-shell gauge
invariance. In $\Lambda=0$ gauge, which reduces to the original unextended
theory, our results agree with those found by Johnson and Sudarshan, and later
verified by Velo and Zwanziger, which give a canonical Rarita-Schwinger field
Dirac bracket that is singular for small gauge fields. In gauge covariant
radiation gauge, the Dirac bracket of the Rarita-Schwinger fields is
nonsingular, but does not correspond to a positive semi-definite
anticommutator, and the Dirac bracket of the auxiliary fields has a singularity
of the same form as found in the unextended theory. These results indicate that
gauged Rarita-Schwinger theory is somewhat pathological, and cannot be
canonically quantized within a conventional positive semi-definite metric
Hilbert space. We leave open the questions of whether consistent quantizations
can be achieved by using an indefinite metric Hilbert space, by path integral
methods, or by appropriate couplings to conventional dimension $\frac{3}{2}$
spin-$\frac{1}{2}$ fields.
| arxiv topic:hep-th |
arxiv_dataset-87871708.03688 | A Statistical Survey of Peculiar L and T Dwarfs in SDSS, 2MASS, and WISE
astro-ph.SR
We present the final results from a targeted search for brown dwarfs with
unusual near-infrared colors. From a positional cross-match of SDSS, 2MASS and
WISE, we have identified 144 candidate peculiar L and T dwarfs. Spectroscopy
confirms that 20 of the objects are peculiar or are candidate binaries. Nine of
the 420 objects in our sample are young ($\lesssim$200 Myr; 2.1%) and another 8
(1.9%) are unusually red with no signatures of youth. With a spectroscopic
$J-K_s$ color of 2.58 $\pm$ 0.11 mag, one of the new objects, the L6 dwarf
2MASS J03530419+0418193, is among the reddest field dwarfs currently known and
is one of the reddest objects with no signatures of youth known to date. We
have also discovered another potentially very low gravity object, the L1 dwarf
2MASS J00133470+1109403, and independently identified the young L7 dwarf 2MASS
J00440332+0228112, first reported by Schneider and collaborators. Our results
confirm that signatures of low gravity are no longer discernible in low to
moderate resolution spectra of objects older than $\sim$200 Myr. The 1.9% of
unusually red L dwarfs that do not show other signatures of youth could be
slightly older, up to $\sim$400 Myr. In this case a red $J-K_s$ color may be
more diagnostic of moderate youth than individual spectral features. However,
its is also possible that these objects are relatively metal-rich, and so have
an enhanced atmospheric dust content.
| arxiv topic:astro-ph.SR |
arxiv_dataset-87881708.03788 | Direct-Manipulation Visualization of Deep Networks
cs.LG cs.HC stat.ML
The recent successes of deep learning have led to a wave of interest from
non-experts. Gaining an understanding of this technology, however, is
difficult. While the theory is important, it is also helpful for novices to
develop an intuitive feel for the effect of different hyperparameters and
structural variations. We describe TensorFlow Playground, an interactive, open
sourced visualization that allows users to experiment via direct manipulation
rather than coding, enabling them to quickly build an intuition about neural
nets.
| arxiv topic:cs.LG cs.HC stat.ML |
arxiv_dataset-87891708.03888 | Large Batch Training of Convolutional Networks
cs.CV
A common way to speed up training of large convolutional networks is to add
computational units. Training is then performed using data-parallel synchronous
Stochastic Gradient Descent (SGD) with mini-batch divided between computational
units. With an increase in the number of nodes, the batch size grows. But
training with large batch size often results in the lower model accuracy. We
argue that the current recipe for large batch training (linear learning rate
scaling with warm-up) is not general enough and training may diverge. To
overcome this optimization difficulties we propose a new training algorithm
based on Layer-wise Adaptive Rate Scaling (LARS). Using LARS, we scaled Alexnet
up to a batch size of 8K, and Resnet-50 to a batch size of 32K without loss in
accuracy.
| arxiv topic:cs.CV |
arxiv_dataset-87901708.03988 | Optimizing the optical imaging system by \emph{in-situ} imaging the
plugged hole in the ultracold atoms
physics.atom-ph physics.optics
Optical absorption imaging has become a common technique for detecting the
density distribution of ultracold atoms. The defocus effect generally produces
artificial spatial structures in the obtained images, which confuses our
understanding of the quantum systems. Here we experimentally demonstrate one
method to optimize the optical imaging system by \emph{in-situ} imaging the
plugged hole in the cold atoms. The atoms confined in a magnetic trap are
cooled to tens of or several microkelvin by the radio-frequency evaporation
cooling, and then are plugged using a blue-detuned laser beam, forming a hole
in the center of the atomic cloud. We image the hole with a charge-coupled
device (CCD) and quantitatively analyze the artificial spatial structure due to
the defocus effect. Through minimizing the artificial structures by precisely
adjusting the CCD position, we can optimize the imaging system with an accuracy
of 0.1 mm. We also demonstrate the necessity of this method in probing rubidium
BEC with a time of flight (TOF) of 5 ms. Compared to other methods in focusing
the imaging system, the proposal demonstrated in this paper is simple and
efficient, particularly for experimentally extracting large-scale parameters
like atomic density, atomic number and the size of the atomic cloud.
| arxiv topic:physics.atom-ph physics.optics |
arxiv_dataset-87911708.04088 | State transfer with quantum side information
quant-ph
We first consider quantum communication protocols between a sender Alice and
a receiver Bob, which transfer Alice's quantum information to Bob by means of
non-local resources, such as classical communication, quantum communication,
and entanglement. In these protocols, we assume that Alice and Bob may have
quantum side information, not transferred. In this work, these protocols are
called the state transfer with quantum side information. We determine the
optimal costs for non-local resources in the protocols, and study what the
effects of the use of quantum side information are. Our results can give new
operational meanings to the quantum mutual information and the quantum
conditional mutual information, which directly provide us with an operational
interpretation of the chain rule for the quantum mutual information.
| arxiv topic:quant-ph |
arxiv_dataset-87921708.04188 | Search for resonant and nonresonant Higgs boson pair production in the
bblnulnu final state in proton-proton collisions at sqrt(s) = 13 TeV
hep-ex
Searches for resonant and nonresonant pair-produced Higgs bosons (HH)
decaying respectively into ll nu nu, through either W or Z bosons, and bbbar
are presented. The analyses are based on a sample of proton-proton collisions
at sqrt(s) = 13 TeV, collected by the CMS experiment at the LHC, corresponding
to an integrated luminosity of 35.9 inverse femtobarns. Data and predictions
from the standard model are in agreement within uncertainties. For the standard
model HH hypothesis, the data exclude at 95% confidence level a product of the
production cross section and branching fraction larger than 72 fb,
corresponding to 79 times the prediction, consistent with expectations.
Constraints are placed on different scenarios considering anomalous couplings,
which could affect the rate and kinematics of HH production. Upper limits at
95% confidence level are set on the production cross section of narrow-width
spin-0 and spin-2 particles decaying to Higgs boson pairs, the latter produced
with minimal gravity-like coupling.
| arxiv topic:hep-ex |
arxiv_dataset-87931708.04288 | Primitive root biases for prime pairs I: existence and non-totality of
biases
math.NT
We study the difference between the number of primitive roots modulo $p$ and
modulo $p+k$ for prime pairs $p,p+k$. Assuming the Bateman-Horn conjecture, we
prove the existence of strong sign biases for such pairs. More importantly, we
prove that for a small positive proportion of prime pairs $p,p+k$, the dominant
inequality is reversed.
| arxiv topic:math.NT |
arxiv_dataset-87941708.04388 | Impact of impurities on zonal flow driven by trapped electron mode
turbulence
physics.plasm-ph
The impact of impurities on the generation of zonal flow (ZF) driven by
collisonless trapped electron mode (CTEM) turbulence in deuterium (D)-tritium
(T) plasmas is investigated. The expression for ZF growth rate with impurities
is derived by balancing the ZF potential shielded by polarization effects and
the ZF modulated radial turbulent current. Then, it is shown that the maximum
normalized ZF growth rate is reduced by the presence of the fully ionized
non-trace light impurities with relatively flat density profile, and slightly
reduced by highly ionized trace tungsten (W). While, the maximum normalized ZF
growth rate can be also enhanced by fully ionized non-trace light impurities
with relatively steep density profile. In particular, the effects of high
temperature helium from D-T reaction on ZF depend on the temperature ratio
between electron and high temperature helium. The possible relevance of our
findings to recent experimental results and future burning plasmas is also
discussed.
| arxiv topic:physics.plasm-ph |
arxiv_dataset-87951708.04488 | Edge-magic labelings for constellations and armies of caterpillars
math.CO
Let $G=(V,E)$ be an $n$-vertex graph with $m$ edges. A function $f : V \cup E
\rightarrow \{1, \ldots, n+m\}$ is an edge-magic labeling of $G$ if $f$ is
bijective and, for some integer $k$, we have $f(u)+f(v)+f(uv) = k$ for every
edge $uv \in E$. Furthermore, if $f(V) = \{1, \ldots, n\}$, then we say that
$f$ is a super edge-magic labeling. A constellation, which is a collection of
stars, is symmetric if the number of stars of each size is even except for at
most one size. We prove that every symmetric constellation with an odd number
of stars admits a super edge-magic labeling. We say that a caterpillar is of
type $(r,s)$ if $r$ and $s$ are the sizes of its parts, where $r \leq s$. We
also prove that every collection with an odd number of same-type caterpillars
admits an edge-magic labeling.
| arxiv topic:math.CO |
arxiv_dataset-87961708.04588 | Stability of spherically symmetric timelike thin-shells in general
relativity with a variable equation of state
gr-qc
We study spherically symmetric timelike thin-shells in $3+1-$dimensional bulk
spacetime with a variable equation of state for the fluid presented on the
shell. In such a fluid the angular pressure $p$ is a function of both surface
energy density $\sigma $ and the radius $R$ of the thin-shell. Explicit cases
of the thin shells connecting two non-identical cloud of strings spacetimes and
a flat Minkowski spacetime to the Schwarzschild metric are investigated.
| arxiv topic:gr-qc |
arxiv_dataset-87971708.04688 | SDSS-IV MaStar: a Large, Comprehensive, and High Quality Empirical
Stellar Library
astro-ph.GA
We introduce the ongoing MaStar project, which is going to construct a large,
well-calibrated, high quality empirical stellar library with more than 8000
stars covering the wavelength range from 3622 to 10,354A at a resolution of
R~2000, and with better than 3% relative flux calibration. The spectra are
taken using hexagonal fiber bundles feeding the BOSS spectrographs on the 2.5m
Sloan Foundation Telescope, by piggybacking on the SDSS-IV/APOGEE-2
observations. Compared to previous efforts of empirical libraries, the MaStar
Library will have a more comprehensive stellar parameter coverage, especially
in cool dwarfs, low metallicity stars, and stars with different [alpha/Fe].
This is achieved by a target selection method based on large spectroscopic
catalogs from APOGEE, LAMOST, and SEGUE, combined with photometric selection.
This empirical library will provide a new basis for calibrating theoretical
spectral libraries and for stellar population synthesis. In addition, with
identical spectral coverage and resolution to the ongoing integral field
spectroscopy survey of nearby galaxies --- SDSS-IV/MaNGA (Mapping Nearby
Galaxies at APO). This library is ideal for spectral modeling and stellar
population analysis of MaNGA data.
| arxiv topic:astro-ph.GA |
arxiv_dataset-87981708.04788 | BitNet: Bit-Regularized Deep Neural Networks
cs.LG stat.ML
We present a novel optimization strategy for training neural networks which
we call "BitNet". The parameters of neural networks are usually unconstrained
and have a dynamic range dispersed over all real values. Our key idea is to
limit the expressive power of the network by dynamically controlling the range
and set of values that the parameters can take. We formulate this idea using a
novel end-to-end approach that circumvents the discrete parameter space by
optimizing a relaxed continuous and differentiable upper bound of the typical
classification loss function. The approach can be interpreted as a
regularization inspired by the Minimum Description Length (MDL) principle. For
each layer of the network, our approach optimizes real-valued translation and
scaling factors and arbitrary precision integer-valued parameters (weights). We
empirically compare BitNet to an equivalent unregularized model on the MNIST
and CIFAR-10 datasets. We show that BitNet converges faster to a superior
quality solution. Additionally, the resulting model has significant savings in
memory due to the use of integer-valued parameters.
| arxiv topic:cs.LG stat.ML |
arxiv_dataset-87991708.04888 | Baryon number fluctuations and QCD phase structure
hep-ph
We investigate the phase structure of strongly interacting matter and baryon
number fluctuations in the Polyakov loop improved Nambu--Jona-Lasinio (PNJL)
model. The calculation shows that both the chiral and deconfinement
transitions, as well as their coincidence and separation determine the basic
QCD phase structure. The contour maps and the three-dimensional diagrams of the
net-baryon kurtosis and skewness present well the trace of QCD phase structure.
Comparing with the experimental data, we find that the existence of a critical
end point (CEP) of chiral transition is crucial to explain the non-monotonic
energy dependence and the large deviation from Poisson baseline of net-proton
kurtosis. In particular, the relation between the chiral and deconfinement
transitions in the crossover region is also reflected by the baryon number
fluctuations. This study shows that the measurements of higher moments of
multiplicity distributions of conserved charges are powerful to investigate the
criticality and even the chiral and deconfinement transitions in the crossover
region.
| arxiv topic:hep-ph |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.