text
null | inputs
dict | prediction
null | prediction_agent
null | annotation
list | annotation_agent
null | multi_label
bool 1
class | explanation
null | id
stringlengths 1
5
| metadata
null | status
stringclasses 2
values | event_timestamp
null | metrics
null |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
null |
{
"abstract": " Despite the growing prominence of generative adversarial networks (GANs),\noptimization in GANs is still a poorly understood topic. In this paper, we\nanalyze the \"gradient descent\" form of GAN optimization i.e., the natural\nsetting where we simultaneously take small gradient steps in both generator and\ndiscriminator parameters. We show that even though GAN optimization does not\ncorrespond to a convex-concave game (even for simple parameterizations), under\nproper conditions, equilibrium points of this optimization procedure are still\n\\emph{locally asymptotically stable} for the traditional GAN formulation. On\nthe other hand, we show that the recently proposed Wasserstein GAN can have\nnon-convergent limit cycles near equilibrium. Motivated by this stability\nanalysis, we propose an additional regularization term for gradient descent GAN\nupdates, which \\emph{is} able to guarantee local stability for both the WGAN\nand the traditional GAN, and also shows practical promise in speeding up\nconvergence and addressing mode collapse.\n",
"title": "Gradient descent GAN optimization is locally stable"
}
| null | null | null | null | true | null |
6101
| null |
Default
| null | null |
null |
{
"abstract": " We tabulate spontaneous emission rates for all possible 811\nelectric-dipole-allowed transitions between the 75 lowest-energy states of Ca\nI. These involve the $4sns$ ($n=4-8$), $4snp$ ($n=4-7$), $4snd$ ($n=3-6$),\n$4snf$ ($n=4-6$), $4p^2$, and $3d4p$ electronic configurations. We compile the\ntransition rates by carrying out ab initio relativistic calculations using the\ncombined method of configuration interaction and many-body perturbation theory.\nThe results are compared to the available literature values. The tabulated\nrates can be useful in various applications, such as optimizing laser cooling\nin magneto-optical traps, estimating various systematic effects in optical\nclocks and evaluating static or dynamic polarizabilities and long-range\natom-atom interaction coefficients and related atomic properties.\n",
"title": "Transition rates and radiative lifetimes of Ca I"
}
| null | null | null | null | true | null |
6102
| null |
Default
| null | null |
null |
{
"abstract": " Various experimental techniques, have revealed that the predominant intrinsic\npoint defects in BaF$_2$ are anion Frenkel defects. Their formation enthalpy\nand entropy as well as the corresponding parameters for the fluorine vacancy\nand fluorine interstitial motion have been determined. In addition, low\ntemperature dielectric relaxation measurements in BaF$_2$ doped with uranium\nleads to the parameters {\\tau}$_0$, E in the Arrhenius relation\n{\\tau}={\\tau}$_0$exp(E/kBT) for the relaxation time {\\tau}. For the relaxation\npeak associated with a single tetravalent uranium, the migration entropy\ndeduced from the pre-exponential factor {\\tau}$_0$, is smaller than the anion\nFrenkel defect formation entropy by almost two orders of magnitude. We show\nthat, despite their great variation, the defect entropies and enthalpies are\ninterconnected through a model based on anharmonic properties of the bulk\nmaterial that have been recently studied by employing density-functional theory\nand density-functional perturbation theory.\n",
"title": "Defect entropies and enthalpies in Barium Fluoride"
}
| null | null | null | null | true | null |
6103
| null |
Default
| null | null |
null |
{
"abstract": " Exploration in complex domains is a key challenge in reinforcement learning,\nespecially for tasks with very sparse rewards. Recent successes in deep\nreinforcement learning have been achieved mostly using simple heuristic\nexploration strategies such as $\\epsilon$-greedy action selection or Gaussian\ncontrol noise, but there are many tasks where these methods are insufficient to\nmake any learning progress. Here, we consider more complex heuristics:\nefficient and scalable exploration strategies that maximize a notion of an\nagent's surprise about its experiences via intrinsic motivation. We propose to\nlearn a model of the MDP transition probabilities concurrently with the policy,\nand to form intrinsic rewards that approximate the KL-divergence of the true\ntransition probabilities from the learned model. One of our approximations\nresults in using surprisal as intrinsic motivation, while the other gives the\n$k$-step learning progress. We show that our incentives enable agents to\nsucceed in a wide range of environments with high-dimensional state spaces and\nvery sparse rewards, including continuous control tasks and games in the Atari\nRAM domain, outperforming several other heuristic exploration techniques.\n",
"title": "Surprise-Based Intrinsic Motivation for Deep Reinforcement Learning"
}
| null | null | null | null | true | null |
6104
| null |
Default
| null | null |
null |
{
"abstract": " We present an experimental study of the local and collective magnetism of\n$\\mathrm{EuFe_2As_2}$, that is isostructural with the high temperature\nsuperconductor parent compound $\\mathrm{BaFe_2As_2}$. In contrast to\n$\\mathrm{BaFe_2As_2}$, where only Fe spins order, $\\mathrm{EuFe_2As_2}$ has an\nadditional magnetic transition below 20 K due to the ordering of the Eu$^{2+}$\nspins ($J =7/2$, with $L=0$ and $S=7/2$) in an A-type antiferromagnetic texture\n(ferromagnetic layers stacked antiferromagnetically). This may potentially\naffect the FeAs layer and its local and correlated magnetism. Fe-K$_\\beta$\nx-ray emission experiments on $\\mathrm{EuFe_2As_2}$ single crystals reveal a\nlocal magnetic moment of 1.3$\\pm0.15~\\mu_B$ at 15 K that slightly increases to\n1.45$\\pm0.15~\\mu_B$ at 300 K. Resonant inelastic x-ray scattering (RIXS)\nexperiments performed on the same crystals show dispersive broad (in energy)\nmagnetic excitations along $\\mathrm{(0, 0)\\rightarrow(1, 0)}$ and $\\mathrm{(0,\n0)\\rightarrow(1, 1)}$ with a bandwidth on the order of 170-180 meV. These\nresults on local and collective magnetism are in line with other parent\ncompounds of the $\\mathrm{AFe_2As_2}$ series ($A=$ Ba, Ca, and Sr), especially\nthe well characterized $\\mathrm{BaFe_2As_2}$. Thus, our experiments lead us to\nthe conclusion that the effect of the high magnetic moment of Eu on the\nmagnitude of both Fe local magnetic moment and spin excitations is small and\nconfined to low energy excitations.\n",
"title": "Local and collective magnetism of EuFe$_2$As$_2$"
}
| null | null |
[
"Physics"
] | null | true | null |
6105
| null |
Validated
| null | null |
null |
{
"abstract": " We fix a monic polynomial $f(x) \\in \\mathbb F_q[x]$ over a finite field of\ncharacteristic $p$ of degree relatively prime to $p$. Let $a\\mapsto \\omega(a)$\nbe the Teichmüller lift of $\\mathbb F_q$, and let $\\chi:\\mathbb{Z}\\to \\mathbb\nC_p^\\times$ be a finite character of $\\mathbb Z_p$. The $L$-function associated\nto the polynomial $f$ and the so-called twisted character $\\omega^u\\times \\chi$\nis denoted by $L_f(\\omega^u,\\chi,s)$. We prove that, when the conductor of the\ncharacter is large enough, the $p$-adic Newton slopes of this $L$-function form\narithmetic progressions.\n",
"title": "Newton slopes for twisted Artin--Schreier--Witt Towers"
}
| null | null | null | null | true | null |
6106
| null |
Default
| null | null |
null |
{
"abstract": " The objective of this work is to augment the basic abilities of a robot by\nlearning to use new sensorimotor primitives to enable the solution of complex\nlong-horizon problems. Solving long-horizon problems in complex domains\nrequires flexible generative planning that can combine primitive abilities in\nnovel combinations to solve problems as they arise in the world. In order to\nplan to combine primitive actions, we must have models of the preconditions and\neffects of those actions: under what circumstances will executing this\nprimitive achieve some particular effect in the world?\nWe use, and develop novel improvements on, state-of-the-art methods for\nactive learning and sampling. We use Gaussian process methods for learning the\nconditions of operator effectiveness from small numbers of expensive training\nexamples collected by experimentation on a robot. We develop adaptive sampling\nmethods for generating diverse elements of continuous sets (such as robot\nconfigurations and object poses) during planning for solving a new task, so\nthat planning is as efficient as possible. We demonstrate these methods in an\nintegrated system, combining newly learned models with an efficient\ncontinuous-space robot task and motion planner to learn to solve long horizon\nproblems more efficiently than was previously possible.\n",
"title": "Active model learning and diverse action sampling for task and motion planning"
}
| null | null | null | null | true | null |
6107
| null |
Default
| null | null |
null |
{
"abstract": " We present a computational method to evaluate the end-to-end and the contour\nlength distribution functions of short DNA molecules described by a mesoscopic\nHamiltonian. The method generates a large statistical ensemble of possible\nconfigurations for each dimer in the sequence, selects the global equilibrium\ntwist conformation for the molecule and determines the average base pair\ndistances along the molecule backbone. Integrating over the base pair radial\nand angular fluctuations, we derive the room temperature distribution functions\nas a function of the sequence length. The obtained values for the most probable\nend-to-end distance and contour length distance, providing a measure of the\nglobal molecule size, are used to examine the DNA flexibility at short length\nscales. It is found that, also in molecules with less than $\\sim 60$ base\npairs, coiled configurations maintain a large statistical weight and,\nconsistently, the persistence lengths may be much smaller than in kilo-base\nDNA.\n",
"title": "End-to-end distance and contour length distribution functions of DNA helices"
}
| null | null | null | null | true | null |
6108
| null |
Default
| null | null |
null |
{
"abstract": " Many time-series data including text, movies, and biological signals can be\nrepresented as sequences of correlated binary patterns. These patterns may be\ndescribed by weighted combinations of a few dominant structures that underpin\nspecific interactions among the binary elements. To extract the dominant\ncorrelation structures and their contributions to generating data in a\ntime-dependent manner, we model the dynamics of binary patterns using the\nstate-space model of an Ising-type network that is composed of multiple\nundirected graphs. We provide a sequential Bayes algorithm to estimate the\ndynamics of weights on the graphs while gaining the graph structures online.\nThis model can uncover overlapping graphs underlying the data better than a\ntraditional orthogonal decomposition method, and outperforms an original\ntime-dependent full Ising model. We assess the performance of the method by\nsimulated data, and demonstrate that spontaneous activity of cultured\nhippocampal neurons is represented by dynamics of multiple graphs.\n",
"title": "Online Estimation of Multiple Dynamic Graphs in Pattern Sequences"
}
| null | null | null | null | true | null |
6109
| null |
Default
| null | null |
null |
{
"abstract": " We introduce a community detection algorithm (Fluid Communities) based on the\nidea of fluids interacting in an environment, expanding and contracting as a\nresult of that interaction. Fluid Communities is based on the propagation\nmethodology, which represents the state-of-the-art in terms of computational\ncost and scalability. While being highly efficient, Fluid Communities is able\nto find communities in synthetic graphs with an accuracy close to the current\nbest alternatives. Additionally, Fluid Communities is the first\npropagation-based algorithm capable of identifying a variable number of\ncommunities in network. To illustrate the relevance of the algorithm, we\nevaluate the diversity of the communities found by Fluid Communities, and find\nthem to be significantly different from the ones found by alternative methods.\n",
"title": "Fluid Communities: A Competitive, Scalable and Diverse Community Detection Algorithm"
}
| null | null | null | null | true | null |
6110
| null |
Default
| null | null |
null |
{
"abstract": " This paper aims to identify three electrical systems: a series RLC circuit, a\nmotor/generator coupled system, and the Duffing-Ueda oscillator. In order to\nobtain the system's models was used the error reduction ratio and the Akaike\ninformation criterion. Our approach to handle the numerical errors was the\ninterval arithmetic by means of the resolution of the least squares estimation.\nThe routines was implemented in Intlab, a Matlab toolbox devoted to arithmetic\ninterval. Finally, the interval RMSE was calculated to verify the quality of\nthe obtained models. The applied methodology was satisfactory, since the\nobtained intervals encompass the system's data and allow to demonstrate how the\nnumerical errors affect the answers.\n",
"title": "Identification of Dynamic Systems with Interval Arithmetic"
}
| null | null | null | null | true | null |
6111
| null |
Default
| null | null |
null |
{
"abstract": " This study has the purpose of addressing four questions that lie at the base\nof the probability theory and statistics, and includes two main steps. As\nfirst, we conduct the textual analysis of the most significant works written by\neminent probability theorists. The textual analysis turns out to be a rather\ninnovative method of study in this domain, and shows how the sampled writers,\nno matter he is a frequentist or a subjectivist, share a similar approach. Each\nauthor argues on the multifold aspects of probability then he establishes the\nmathematical theory on the basis of his intellectual conclusions. It may be\nsaid that mathematics ranks second. Hilbert foresees an approach far different\nfrom that used by the sampled authors. He proposes to axiomatize the\nprobability calculus notably to describe the probability concepts using purely\nmathematical criteria. In the second stage of the present research we address\nthe four issues of the probability theory and statistics following the\nrecommendations of Hilbert. Specifically, we use two theorems that prove how\nthe frequentist and the subjectivist models are not incompatible as many\nbelieve. Probability has distinct meanings under different hypotheses, and in\nturn classical statistics and Bayesian statistics are available for adoption in\ndifferent circumstances. Subsequently, these results are commented upon,\nfollowed by our conclusions\n",
"title": "Four Fundamental Questions in Probability Theory and Statistics"
}
| null | null | null | null | true | null |
6112
| null |
Default
| null | null |
null |
{
"abstract": " We define parahoric $\\cG$--torsors for certain Bruhat--Tits group scheme\n$\\cG$ on a smooth complex projective curve $X$ when the weights are real, and\nalso define connections on them. We prove that a $\\cG$--torsor is given by a\nhomomorphism from $\\pi_1(X\\setminus D)$ to a maximal compact subgroup of $G$,\nwhere $D\\, \\subset\\, X$ is the parabolic divisor, if and only if the torsor is\npolystable.\n",
"title": "Connections on parahoric torsors over curves"
}
| null | null | null | null | true | null |
6113
| null |
Default
| null | null |
null |
{
"abstract": " The identification of reduced-order models from high-dimensional data is a\nchallenging task, and even more so if the identified system should not only be\nsuitable for a certain data set, but generally approximate the input-output\nbehavior of the data source. In this work, we consider the input-output dynamic\nmode decomposition method for system identification. We compare excitation\napproaches for the data-driven identification process and describe an\noptimization-based stabilization strategy for the identified systems.\n",
"title": "On Reduced Input-Output Dynamic Mode Decomposition"
}
| null | null | null | null | true | null |
6114
| null |
Default
| null | null |
null |
{
"abstract": " Apprenticeship learning (AL) is a kind of Learning from Demonstration\ntechniques where the reward function of a Markov Decision Process (MDP) is\nunknown to the learning agent and the agent has to derive a good policy by\nobserving an expert's demonstrations. In this paper, we study the problem of\nhow to make AL algorithms inherently safe while still meeting its learning\nobjective. We consider a setting where the unknown reward function is assumed\nto be a linear combination of a set of state features, and the safety property\nis specified in Probabilistic Computation Tree Logic (PCTL). By embedding\nprobabilistic model checking inside AL, we propose a novel\ncounterexample-guided approach that can ensure safety while retaining\nperformance of the learnt policy. We demonstrate the effectiveness of our\napproach on several challenging AL scenarios where safety is essential.\n",
"title": "Safety-Aware Apprenticeship Learning"
}
| null | null | null | null | true | null |
6115
| null |
Default
| null | null |
null |
{
"abstract": " Some effects of surface tension on fully-nonlinear, long, surface water waves\nare studied by numerical means. The differences between various solitary waves\nand their interactions in subcritical and supercritical surface tension regimes\nare presented. Analytical expressions for new peaked travelling wave solutions\nare presented in the case of critical surface tension. The numerical\nexperiments were performed using a high-accurate finite element method based on\nsmooth cubic splines and the four-stage, classical, explicit Runge-Kutta method\nof order four.\n",
"title": "Solitary wave solutions and their interactions for fully nonlinear water waves with surface tension in the generalized Serre equations"
}
| null | null | null | null | true | null |
6116
| null |
Default
| null | null |
null |
{
"abstract": " A characterization of relative weak mixing in W*-dynamical systems in terms\nof a relatively independent joining is proven.\n",
"title": "Relative weak mixing of W*-dynamical systems via joinings"
}
| null | null | null | null | true | null |
6117
| null |
Default
| null | null |
null |
{
"abstract": " The Extreme Ultraviolet Variability Experiment (EVE) on the Solar Dynamics\nObservatory obtains extreme-ultraviolet (EUV) spectra of the full-disk Sun at a\nspectral resolution of ~1 A and cadence of 10 s. Such a spectral resolution\nwould normally be considered to be too low for the reliable determination of\nelectron density (N_e) sensitive emission line intensity ratios, due to\nblending. However, previous work has shown that a limited number of Fe XXI\nfeatures in the 90-60 A wavelength region of EVE do provide useful\nN_e-diagnostics at relatively low flare densities (N_e ~ 10^11-10^12 cm^-3).\nHere we investigate if additional highly ionised Fe line ratios in the EVE\n90-160 A range may be reliably employed as N_e-diagnostics. In particular, the\npotential for such diagnostics to provide density estimates for high N_e\n(~10^13 cm^-3) flare plasmas is assessed. Our study employs EVE spectra for\nX-class flares, combined with observations of highly active late-type stars\nfrom the Extreme Ultraviolet Explorer (EUVE) satellite plus experimental data\nfor well-diagnosed tokamak plasmas, both of which are similar in wavelength\ncoverage and spectral resolution to those from EVE. Several ratios are\nidentified in EVE data which yield consistent values of electron density,\nincluding Fe XX 113.35/121.85 and Fe XXII 114.41/135.79, with confidence in\ntheir reliability as N_e-diagnostics provided by the EUVE and tokamak results.\nThese ratios also allow the determination of density in solar flare plasmas up\nto values of ~10^13 cm^-3.\n",
"title": "An assessment of Fe XX - Fe XXII emission lines in SDO/EVE data as diagnostics for high density solar flare plasmas using EUVE stellar observations"
}
| null | null | null | null | true | null |
6118
| null |
Default
| null | null |
null |
{
"abstract": " Weighting pixel contribution considering its location is a key feature in\nmany fundamental image processing tasks including filtering, object modeling\nand distance matching. Several techniques have been proposed that incorporate\nSpatial information to increase the accuracy and boost the performance of\ndetection, tracking and recognition systems at the cost of speed. But, it is\nstill not clear how to efficiently ex- tract weighted local histograms in\nconstant time using integral histogram. This paper presents a novel algorithm\nto compute accurately multi-scale Spatially weighted local histograms in\nconstant time using Weighted Integral Histogram (SWIH) for fast search. We\napplied our spatially weighted integral histogram approach for fast tracking\nand obtained more accurate and robust target localization result in comparison\nwith using plain histogram.\n",
"title": "Multi-Scale Spatially Weighted Local Histograms in O(1)"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6119
| null |
Validated
| null | null |
null |
{
"abstract": " We study the dispersion of a point set, a notion closely related to the\ndiscrepancy. Given a real $r\\in (0,1)$ and an integer $d\\geq 2$, let $N(r,d)$\ndenote the minimum number of points inside the $d$-dimensional unit cube\n$[0,1]^d$ such that they intersect every axis-aligned box inside $[0,1]^d$ of\nvolume greater than $r$. We prove an upper bound on $N(r,d)$, matching a lower\nbound of Aistleitner et al. up to a multiplicative constant depending only on\n$r$. This fully determines the rate of growth of $N(r,d)$ if $r\\in(0,1)$ is\nfixed.\n",
"title": "A note on minimal dispersion of point sets in the unit cube"
}
| null | null | null | null | true | null |
6120
| null |
Default
| null | null |
null |
{
"abstract": " In this paper additive bi-free convolution is defined for general Borel\nprobability measures, and the limiting distributions for sums of bi-free pairs\nof selfadjoint commuting random variables in an infinitesimal triangular array\nare determined. These distributions are characterized by their bi-freely\ninfinite divisibility, and moreover, a transfer principle is established for\nlimit theorems in classical probability theory and Voiculescu's bi-free\nprobability theory. Complete descriptions of bi-free stability and fullness of\nplanar probability distributions are also set down. All these results reveal\none important feature about the theory of bi-free probability that it parallels\nthe classical theory perfectly well. The emphasis in the whole work is not on\nthe tool of bi-free combinatorics but only on the analytic machinery.\n",
"title": "Limit theorems in bi-free probability theory"
}
| null | null | null | null | true | null |
6121
| null |
Default
| null | null |
null |
{
"abstract": " We characterise finite axiomatisability and intractability of deciding\nmembership for universal Horn classes generated by finite loop-free\nhypergraphs.\n",
"title": "Axiomatisability and hardness for universal Horn classes of hypergraphs"
}
| null | null | null | null | true | null |
6122
| null |
Default
| null | null |
null |
{
"abstract": " This work is focused on searching a geodesic interpretation of the dynamics\nof a particle under the effects of a Snyder like deformation in the background\nof the Kepler problem. In order to accomplish that task, a newtonian spacetime\nis used. Newtonian spacetime is not a metric manifold, but allows to introduce\na torsion free connection in order to interpret the dynamic equations of the\ndeformed Kepler problem as geodesics in a curved spacetime. These geodesics and\nthe curvature terms of the Riemann and Ricci tensors show a mass and a\nfundamental length dependence as expected, but are velocity independent. In\nthis sense, the effect of introducing a deformed algebra is examinated and the\ncorresponding curvature terms calculated, as well as the modifications of the\nintegrals of motion.\n",
"title": "Snyder Like Modified Gravity in Newton's Spacetime"
}
| null | null | null | null | true | null |
6123
| null |
Default
| null | null |
null |
{
"abstract": " A photonic circuit is generally described as a structure in which light\npropagates by unitary exchange and transfers reversibly between channels. In\ncontrast, the term `diffusive' is more akin to a chaotic propagation in\nscattering media, where light is driven out of coherence towards a thermal\nmixture. Based on the dynamics of open quantum systems, the combination of\nthese two opposites can result in novel techniques for coherent light control.\nThe crucial feature of these photonic structures is dissipative coupling\nbetween modes, via an interaction with a common reservoir. Here, we demonstrate\nexperimentally that such systems can perform optical equalisation to smooth\nmultimode light, or act as a distributor, guiding it into selected channels.\nQuantum thermodynamically, these systems can act as catalytic coherent\nreservoirs by performing perfect non-Landauer erasure. For lattice structures,\nlocalised stationary states can be supported in the continuum, similar to\ncompacton-like states in conventional flat band lattices.\n",
"title": "Dissipatively Coupled Waveguide Networks for Coherent Diffusive Photonics"
}
| null | null | null | null | true | null |
6124
| null |
Default
| null | null |
null |
{
"abstract": " In the early 90s, researchers began to focus on security as an important\nproperty to address in combination with safety. Over the years, researchers\nhave proposed approaches to harmonize activities within the safety and security\ndisciplines. Despite the academic efforts to identify interdependencies and to\npropose combined approaches for safety and security, there is still a lack of\nintegration between safety and security practices in the industrial context, as\nthey have separate standards and independent processes often addressed and\nassessed by different organizational teams and authorities. Specifically,\nsecurity concerns are generally not covered in any detail in safety standards\npotentially resulting in successfully safety-certified systems that still are\nopen for security threats from e.g., malicious intents from internal and\nexternal personnel and hackers that may jeopardize safety. In recent years\nsecurity has again received an increasing attention of being an important issue\nalso in safety assurance, as the open interconnected nature of emerging systems\nmakes them susceptible to security threats at a much higher degree than\nexisting more confined products.This article presents initial ideas on how to\nextend safety work to include aspects of security during the context\nestablishment and initial risk assessment procedures. The ambition of our\nproposal is to improve safety and increase efficiency and effectiveness of the\nsafety work within the frames of the current safety standards, i.e., raised\nsecurity awareness in compliance with the current safety standards. We believe\nthat our proposal is useful to raise the security awareness in industrial\ncontexts, although it is not a complete harmonization of safety and security\ndisciplines, as it merely provides applicable guidance to increase security\nawareness in a safety context.\n",
"title": "Inadequate Risk Analysis Might Jeopardize The Functional Safety of Modern Systems"
}
| null | null | null | null | true | null |
6125
| null |
Default
| null | null |
null |
{
"abstract": " In this work, we consider a one-dimensional It{ô} diffusion process X t\nwith possibly nonlinear drift and diffusion coefficients. We show that, when\nthe diffusion coefficient is known, the drift coefficient is uniquely\ndetermined by an observation of the expectation of the process during a small\ntime interval, and starting from values X 0 in a given subset of R. With the\nsame type of observation, and given the drift coefficient, we also show that\nthe diffusion coefficient is uniquely determined. When both coefficients are\nunknown, we show that they are simultaneously uniquely determined by the\nobservation of the expectation and variance of the process, during a small time\ninterval, and starting again from values X 0 in a given subset of R. To derive\nthese results, we apply the Feynman-Kac theorem which leads to a linear\nparabolic equation with unknown coefficients in front of the first and second\norder terms. We then solve the corresponding inverse problem with PDE technics\nwhich are mainly based on the strong parabolic maximum principle.\n",
"title": "Simultaneous determination of the drift and diffusion coefficients in stochastic differential equations"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6126
| null |
Validated
| null | null |
null |
{
"abstract": " Haslhofer and Müller proved a compactness Theorem for four-dimensional\nshrinking gradient Ricci solitons, with the only assumption being that the\nentropy is uniformly bounded from below. However, the limit in their result\ncould possibly be an orbifold Ricci shrinker. In this paper we prove a\ncompactness theorem for noncompact four-dimensional shrinking gradient Ricci\nsolitons with a topological restriction and a noncollapsing assumption, that\nis, we consider Ricci shrinkers that can be embedded in a closed four-manifold\nwith vanishing second homology group over every field and are strongly\n$\\kappa$-noncollapsed with respect to a universal $\\kappa$. In particular, we\ndo not need any curvature assumption and the limit is still a smooth nonflat\nshrinking gradient Ricci soliton.\n",
"title": "A compactness theorem for four-dimensional shrinking gradient Ricci solitons"
}
| null | null | null | null | true | null |
6127
| null |
Default
| null | null |
null |
{
"abstract": " Wardrop equilibria in nonatomic congestion games are in general inefficient\nas they do not induce an optimal flow that minimizes the total travel time.\nNetwork tolls are a prominent and popular way to induce an optimum flow in\nequilibrium. The classical approach to find such tolls is marginal cost pricing\nwhich requires the exact knowledge of the demand on the network. In this paper,\nwe investigate under which conditions demand-independent optimum tolls exist\nthat induce the system optimum flow for any travel demand in the network. We\ngive several characterizations for the existence of such tolls both in terms of\nthe cost structure and the network structure of the game. Specifically we show\nthat demand-independent optimum tolls exist if and only if the edge cost\nfunctions are shifted monomials as used by the Bureau of Public Roads.\nMoreover, non-negative demand-independent optimum tolls exist when the network\nis a directed acyclic multi-graph. Finally, we show that any network with a\nsingle origin-destination pair admits demand-independent optimum tolls that,\nalthough not necessarily non-negative, satisfy a budget constraint.\n",
"title": "Demand-Independent Optimal Tolls"
}
| null | null | null | null | true | null |
6128
| null |
Default
| null | null |
null |
{
"abstract": " We consider the problem of learning a binary classifier from only positive\nand unlabeled observations (PU learning). Although recent research in PU\nlearning has succeeded in showing theoretical and empirical performance, most\nexisting algorithms need to solve either a convex or a non-convex optimization\nproblem and thus are not suitable for large-scale datasets. In this paper, we\npropose a simple yet theoretically grounded PU learning algorithm by extending\nthe previous work proposed for supervised binary classification (Sriperumbudur\net al., 2012). The proposed PU learning algorithm produces a closed-form\nclassifier when the hypothesis space is a closed ball in reproducing kernel\nHilbert space. In addition, we establish upper bounds of the estimation error\nand the excess risk. The obtained estimation error bound is sharper than\nexisting results and the excess risk bound does not rely on an approximation\nerror term. To the best of our knowledge, we are the first to explicitly derive\nthe excess risk bound in the field of PU learning. Finally, we conduct\nextensive numerical experiments using both synthetic and real datasets,\ndemonstrating improved accuracy, scalability, and robustness of the proposed\nalgorithm.\n",
"title": "An analytic formulation for positive-unlabeled learning via weighted integral probability metric"
}
| null | null | null | null | true | null |
6129
| null |
Default
| null | null |
null |
{
"abstract": " Knights Landing (KNL) is the code name for the second-generation Intel Xeon\nPhi product family. KNL has generated significant interest in the data analysis\nand machine learning communities because its new many-core architecture targets\nboth of these workloads. The KNL many-core vector processor design enables it\nto exploit much higher levels of parallelism. At the Lincoln Laboratory\nSupercomputing Center (LLSC), the majority of users are running data analysis\napplications such as MATLAB and Octave. More recently, machine learning\napplications, such as the UC Berkeley Caffe deep learning framework, have\nbecome increasingly important to LLSC users. Thus, the performance of these\napplications on KNL systems is of high interest to LLSC users and the broader\ndata analysis and machine learning communities. Our data analysis benchmarks of\nthese application on the Intel KNL processor indicate that single-core\ndouble-precision generalized matrix multiply (DGEMM) performance on KNL systems\nhas improved by ~3.5x compared to prior Intel Xeon technologies. Our data\nanalysis applications also achieved ~60% of the theoretical peak performance.\nAlso a performance comparison of a machine learning application, Caffe, between\nthe two different Intel CPUs, Xeon E5 v3 and Xeon Phi 7210, demonstrated a 2.7x\nimprovement on a KNL node.\n",
"title": "Benchmarking Data Analysis and Machine Learning Applications on the Intel KNL Many-Core Processor"
}
| null | null | null | null | true | null |
6130
| null |
Default
| null | null |
null |
{
"abstract": " The conjecture of Lehmer is proved to be true. The proof mainly relies upon:\n(i) the properties of the Parry Upper functions $f_{house(\\alpha)}(z)$\nassociated with the dynamical zeta functions $\\zeta_{house(\\alpha)}(z)$ of the\nRényi--Parry arithmetical dynamical systems, for $\\alpha$ an algebraic\ninteger $\\alpha$ of house \"$house(\\alpha)$\" greater than 1, (ii) the discovery\nof lenticuli of poles of $\\zeta_{house(\\alpha)}(z)$ which uniformly\nequidistribute at the limit on a limit \"lenticular\" arc of the unit circle,\nwhen $house(\\alpha)$ tends to $1^+$, giving rise to a continuous lenticular\nminorant ${\\rm M}_{r}(house(\\alpha))$ of the Mahler measure ${\\rm M}(\\alpha)$,\n(iii) the Poincaré asymptotic expansions of these poles and of this minorant\n${\\rm M}_{r}(house(\\alpha))$ as a function of the dynamical degree. With the\nsame arguments the conjecture of Schinzel-Zassenhaus is proved to be true. An\ninequality improving those of Dobrowolski and Voutier ones is obtained. The set\nof Salem numbers is shown to be bounded from below by the Perron number\n$\\theta_{31}^{-1} = 1.08545\\ldots$, dominant root of the trinomial $-1 - z^{30}\n+ z^{31}$. Whether Lehmer's number is the smallest Salem number remains open. A\nlower bound for the Weil height of nonzero totally real algebraic numbers,\n$\\neq \\pm 1$, is obtained (Bogomolov property). For sequences of algebraic\nintegers of Mahler measure smaller than the smallest Pisot number, whose houses\nhave a dynamical degree tending to infinity, the Galois orbit measures of\nconjugates are proved to converge towards the Haar measure on $|z|=1$ (limit\nequidistribution).\n",
"title": "A Proof of the Conjecture of Lehmer and of the Conjecture of Schinzel-Zassenhaus"
}
| null | null | null | null | true | null |
6131
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study the existence and uniqueness of pseudo\n$S$-asymptotically $\\omega$-periodic mild solutions of class $r$ for fractional\nintegro-differential neutral equations. An example is presented to illustrate\nthe application of the abstract results.\n",
"title": "Pseudo asymptotically periodic solutions for fractional integro-differential neutral equations"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6132
| null |
Validated
| null | null |
null |
{
"abstract": " The 2010 Silent Speech Challenge benchmark is updated with new results\nobtained in a Deep Learning strategy, using the same input features and\ndecoding strategy as in the original article. A Word Error Rate of 6.4% is\nobtained, compared to the published value of 17.4%. Additional results\ncomparing new auto-encoder-based features with the original features at reduced\ndimensionality, as well as decoding scenarios on two different language models,\nare also presented. The Silent Speech Challenge archive has been updated to\ncontain both the original and the new auto-encoder features, in addition to the\noriginal raw data.\n",
"title": "Updating the silent speech challenge benchmark with deep learning"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6133
| null |
Validated
| null | null |
null |
{
"abstract": " The magnetic field induced rearrangement of the cycloidal spin structure in\nferroelectric mono-domain single crystals of the room-temperature multiferroic\nBiFeO$_3$ is studied using small-angle neutron scattering (SANS). The cycloid\npropagation vectors are observed to rotate when magnetic fields applied\nperpendicular to the rhombohedral (polar) axis exceed a pinning threshold value\nof $\\sim$5\\,T. In light of these experimental results, a phenomenological model\nis proposed that captures the rearrangement of the cycloidal domains, and we\nrevisit the microscopic origin of the magnetoelectric effect. A new coupling\nbetween the magnetic anisotropy and the polarization is proposed that explains\nthe recently discovered magnetoelectric polarization to the rhombohedral axis.\n",
"title": "Magnetic field control of cycloidal domains and electric polarization in multiferroic BiFeO$_3$"
}
| null | null | null | null | true | null |
6134
| null |
Default
| null | null |
null |
{
"abstract": " We present the latest major release version 6.0 of the quantified Boolean\nformula (QBF) solver DepQBF, which is based on QCDCL. QCDCL is an extension of\nthe conflict-driven clause learning (CDCL) paradigm implemented in state of the\nart propositional satisfiability (SAT) solvers. The Q-resolution calculus\n(QRES) is a QBF proof system which underlies QCDCL. QCDCL solvers can produce\nQRES proofs of QBFs in prenex conjunctive normal form (PCNF) as a byproduct of\nthe solving process. In contrast to traditional QCDCL based on QRES, DepQBF 6.0\nimplements a variant of QCDCL which is based on a generalization of QRES. This\ngeneralization is due to a set of additional axioms and leaves the original\nQ-resolution rules unchanged. The generalization of QRES enables QCDCL to\npotentially produce exponentially shorter proofs than the traditional variant.\nWe present an overview of the features implemented in DepQBF and report on\nexperimental results which demonstrate the effectiveness of generalized QRES in\nQCDCL.\n",
"title": "DepQBF 6.0: A Search-Based QBF Solver Beyond Traditional QCDCL"
}
| null | null | null | null | true | null |
6135
| null |
Default
| null | null |
null |
{
"abstract": " In this contribution we present numerical and experimental results of a\nparametric study of radiative dipole antennas in a phased array configuration\nfor efficient body magnetic resonance imaging at 7T via parallel transmit. For\nmagnetic resonance imaging (MRI) at ultrahigh fields (7T and higher) dipole\nantennas are commonly used in phased arrays, particularly for body imaging\ntargets. This study reveals the effects of dipole positioning in the array\n(elevation of dipoles above the subject and inter-dipole spacing) on their\nmutual coupling, $B_1^{+}$ per unit power and $B_1^{+}$ per maximum local SAR\nefficiencies as well as the RF-shimming capability. The results demonstrate the\ntrade-off between low maximum local SAR and sensitivity to the subject\nvariation and provide the working parameter range for practical body arrays\ncomposed of recently suggested fractionated dipoles.\n",
"title": "On Optimization of Radiative Dipole Body Array Coils for 7 Tesla MRI"
}
| null | null | null | null | true | null |
6136
| null |
Default
| null | null |
null |
{
"abstract": " The importance of speaking style authentication from human speech is gaining\nan increasing attention and concern from the engineering community. The\nimportance comes from the demand to enhance both the naturalness and efficiency\nof spoken language human-machine interface. Our work in this research focuses\non proposing, implementing, and testing speaker-dependent and text-dependent\nspeaking style authentication (verification) systems that accept or reject the\nidentity claim of a speaking style based on suprasegmental hidden Markov models\n(SPHMMs). Based on using SPHMMs, our results show that the average speaking\nstyle authentication performance is: 99%, 37%, 85%, 60%, 61%, 59%, 41%, 61%,\nand 57% belonging respectively to the speaking styles: neutral, shouted, slow,\nloud, soft, fast, angry, happy, and fearful.\n",
"title": "Speaking Style Authentication Using Suprasegmental Hidden Markov Models"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6137
| null |
Validated
| null | null |
null |
{
"abstract": " The main result in this paper is a fixed point formula for equivariant\nindices of elliptic differential operators, for proper actions by connected\nsemisimple Lie groups on possibly noncompact manifolds, with compact quotients.\nFor compact groups and manifolds, this reduces to the Atiyah-Segal-Singer fixed\npoint formula. Other special cases include an index theorem by Connes and\nMoscovici for homogeneous spaces, and an earlier index theorem by the second\nauthor, both in cases where the group acting is connected and semisimple. As an\napplication of this fixed point formula, we give a new proof of\nHarish-Chandra's character formula for discrete series representations.\n",
"title": "A fixed point formula and Harish-Chandra's character formula"
}
| null | null | null | null | true | null |
6138
| null |
Default
| null | null |
null |
{
"abstract": " We study the fundamental question of the lattice dynamics of a metallic\nferromagnet in the regime where the static long range magnetic order is\nreplaced by the fluctuating local moments embedded in a metallic host. We use\nthe \\textit{ab initio} Density Functional Theory(DFT)+embedded Dynamical\nMean-Field Theory(eDMFT) functional approach to address the dynamic stability\nof iron polymorphs and the phonon softening with increased temperature. We show\nthat the non-harmonic and inhomogeneous phonon softening measured in iron is a\nresult of the melting of the long range ferromagnetic order, and is unrelated\nto the first order structural transition from the BCC to the FCC phase, as is\nusually assumed. We predict that the BCC structure is dynamically stable at all\ntemperatures at normal pressure, and is only thermodynamically unstable between\nthe BCC-$\\alpha$ and the BCC-$\\delta$ phase of iron.\n",
"title": "The phonon softening due to melting of the ferromagnetic order in elemental iron"
}
| null | null | null | null | true | null |
6139
| null |
Default
| null | null |
null |
{
"abstract": " The decomposability of a Cartesian product of two nondecomposable manifolds\ninto products of lower dimensional manifolds is studied. For 3-manifolds we\nobtain an analog of a result due to Borsuk for surfaces, and in higher\ndimensions we show that similar analogs do not exist unless one imposes further\nrestrictions such as simple connectivity.\n",
"title": "Decomposing manifolds into Cartesian products"
}
| null | null | null | null | true | null |
6140
| null |
Default
| null | null |
null |
{
"abstract": " MAGIC, a system of two imaging atmospheric Cherenkov telescopes, achieves its\nbest performance under dark conditions, i.e. in absence of moonlight or\ntwilight. Since operating the telescopes only during dark time would severely\nlimit the duty cycle, observations are also performed when the Moon is present\nin the sky. Here we present a dedicated Moon-adapted analysis and characterize\nthe performance of MAGIC under moonlight. We evaluate energy threshold, angular\nresolution and sensitivity of MAGIC under different background light levels,\nbased on Crab Nebula observations and tuned Monte Carlo simulations. This study\nincludes observations taken under non-standard hardware configurations, such as\nreducing the camera photomultiplier tubes gain by a factor $\\sim$1.7 (reduced\nHV settings) with respect to standard settings (nominal HV) or using UV-pass\nfilters to strongly reduce the amount of moonlight reaching the telescopes\ncameras. The Crab Nebula spectrum is correctly reconstructed in all the studied\nillumination levels, that reach up to 30 times brighter than under dark\nconditions. The main effect of moonlight is an increase in the analysis energy\nthreshold and in the systematic uncertainties on the flux normalization. The\nsensitivity degradation is constrained to be below 10\\%, within 15-30\\% and\nbetween 60 and 80\\% for nominal HV, reduced HV and UV-pass filter observations,\nrespectively. No worsening of the angular resolution was found. Thanks to\nobservations during moonlight, the duty cycle can be doubled, suppressing the\nneed to stop observations around full Moon.\n",
"title": "Performance of the MAGIC telescopes under moonlight"
}
| null | null |
[
"Physics"
] | null | true | null |
6141
| null |
Validated
| null | null |
null |
{
"abstract": " Heterogeneity is one important feature of complex systems, leading to the\ncomplexity of their construction and analysis. Moving the heterogeneity at\nmodel level helps in mastering the difficulty of composing heterogeneous models\nwhich constitute a large system. We propose a method made of an algebra and\nstructure morphisms to deal with the interaction of behavioural models,\nprovided that they are compatible. We prove that heterogeneous models can\ninteract in a safe way, and therefore complex heterogeneous systems can be\nbuilt and analysed incrementally. The Uppaal tool is targeted for\nexperimentations.\n",
"title": "Mastering Heterogeneous Behavioural Models"
}
| null | null | null | null | true | null |
6142
| null |
Default
| null | null |
null |
{
"abstract": " Generative adversarial networks (GAN) approximate a target data distribution\nby jointly optimizing an objective function through a \"two-player game\" between\na generator and a discriminator. Despite their empirical success, however, two\nvery basic questions on how well they can approximate the target distribution\nremain unanswered. First, it is not known how restricting the discriminator\nfamily affects the approximation quality. Second, while a number of different\nobjective functions have been proposed, we do not understand when convergence\nto the global minima of the objective function leads to convergence to the\ntarget distribution under various notions of distributional convergence.\nIn this paper, we address these questions in a broad and unified setting by\ndefining a notion of adversarial divergences that includes a number of recently\nproposed objective functions. We show that if the objective function is an\nadversarial divergence with some additional conditions, then using a restricted\ndiscriminator family has a moment-matching effect. Additionally, we show that\nfor objective functions that are strict adversarial divergences, convergence in\nthe objective function implies weak convergence, thus generalizing previous\nresults.\n",
"title": "Approximation and Convergence Properties of Generative Adversarial Learning"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6143
| null |
Validated
| null | null |
null |
{
"abstract": " The World Health Organization (WHO) reported 1.25 million deaths yearly due\nto road traffic accidents worldwide and the number has been continuously\nincreasing over the last few years. Nearly fifth of these accidents are caused\nby distracted drivers. Existing work of distracted driver detection is\nconcerned with a small set of distractions (mostly, cell phone usage).\nUnreliable ad-hoc methods are often used.In this paper, we present the first\npublicly available dataset for driver distraction identification with more\ndistraction postures than existing alternatives. In addition, we propose a\nreliable deep learning-based solution that achieves a 90% accuracy. The system\nconsists of a genetically-weighted ensemble of convolutional neural networks,\nwe show that a weighted ensemble of classifiers using a genetic algorithm\nyields in a better classification confidence. We also study the effect of\ndifferent visual elements in distraction detection by means of face and hand\nlocalizations, and skin segmentation. Finally, we present a thinned version of\nour ensemble that could achieve 84.64% classification accuracy and operate in a\nreal-time environment.\n",
"title": "Driver Distraction Identification with an Ensemble of Convolutional Neural Networks"
}
| null | null | null | null | true | null |
6144
| null |
Default
| null | null |
null |
{
"abstract": " Advanced driver assistance systems (ADAS) can be significantly improved with\neffective driver action prediction (DAP). Predicting driver actions early and\naccurately can help mitigate the effects of potentially unsafe driving\nbehaviors and avoid possible accidents. In this paper, we formulate driver\naction prediction as a timeseries anomaly prediction problem. While the anomaly\n(driver actions of interest) detection might be trivial in this context,\nfinding patterns that consistently precede an anomaly requires searching for or\nextracting features across multi-modal sensory inputs. We present such a driver\naction prediction system, including a real-time data acquisition, processing\nand learning framework for predicting future or impending driver action. The\nproposed system incorporates camera-based knowledge of the driving environment\nand the driver themselves, in addition to traditional vehicle dynamics. It then\nuses a deep bidirectional recurrent neural network (DBRNN) to learn the\ncorrelation between sensory inputs and impending driver behavior achieving\naccurate and high horizon action prediction. The proposed system performs\nbetter than other existing systems on driver action prediction tasks and can\naccurately predict key driver actions including acceleration, braking, lane\nchange and turning at durations of 5sec before the action is executed by the\ndriver.\n",
"title": "Driver Action Prediction Using Deep (Bidirectional) Recurrent Neural Network"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6145
| null |
Validated
| null | null |
null |
{
"abstract": " Relational reasoning is a central component of generally intelligent\nbehavior, but has proven difficult for neural networks to learn. In this paper\nwe describe how to use Relation Networks (RNs) as a simple plug-and-play module\nto solve problems that fundamentally hinge on relational reasoning. We tested\nRN-augmented networks on three tasks: visual question answering using a\nchallenging dataset called CLEVR, on which we achieve state-of-the-art,\nsuper-human performance; text-based question answering using the bAbI suite of\ntasks; and complex reasoning about dynamic physical systems. Then, using a\ncurated dataset called Sort-of-CLEVR we show that powerful convolutional\nnetworks do not have a general capacity to solve relational questions, but can\ngain this capacity when augmented with RNs. Our work shows how a deep learning\narchitecture equipped with an RN module can implicitly discover and learn to\nreason about entities and their relations.\n",
"title": "A simple neural network module for relational reasoning"
}
| null | null | null | null | true | null |
6146
| null |
Default
| null | null |
null |
{
"abstract": " We propose a new algorithm for the computation of a singular value\ndecomposition (SVD) low-rank approximation of a matrix in the Matrix Product\nOperator (MPO) format, also called the Tensor Train Matrix format. Our tensor\nnetwork randomized SVD (TNrSVD) algorithm is an MPO implementation of the\nrandomized SVD algorithm that is able to compute dominant singular values and\ntheir corresponding singular vectors. In contrast to the state-of-the-art\ntensor-based alternating least squares SVD (ALS-SVD) and modified alternating\nleast squares SVD (MALS-SVD) matrix approximation methods, TNrSVD can be up to\n17 times faster while achieving the same accuracy. In addition, our TNrSVD\nalgorithm also produces accurate approximations in particular cases where both\nALS-SVD and MALS-SVD fail to converge. We also propose a new algorithm for the\nfast conversion of a sparse matrix into its corresponding MPO form, which is up\nto 509 times faster than the standard Tensor Train SVD (TT-SVD) method while\nachieving machine precision accuracy. The efficiency and accuracy of both\nalgorithms are demonstrated in numerical experiments.\n",
"title": "Computing low-rank approximations of large-scale matrices with the Tensor Network randomized SVD"
}
| null | null | null | null | true | null |
6147
| null |
Default
| null | null |
null |
{
"abstract": " Capacitated fixed-charge network flows are used to model a variety of\nproblems in telecommunication, facility location, production planning and\nsupply chain management. In this paper, we investigate capacitated path\nsubstructures and derive strong and easy-to-compute \\emph{path cover and path\npack inequalities}. These inequalities are based on an explicit\ncharacterization of the submodular inequalities through a fast computation of\nparametric minimum cuts on a path, and they generalize the well-known flow\ncover and flow pack inequalities for the single-node relaxations of\nfixed-charge flow models. We provide necessary and sufficient facet conditions.\nComputational results demonstrate the effectiveness of the inequalities when\nused as cuts in a branch-and-cut algorithm.\n",
"title": "Path Cover and Path Pack Inequalities for the Capacitated Fixed-Charge Network Flow Problem"
}
| null | null |
[
"Computer Science",
"Mathematics"
] | null | true | null |
6148
| null |
Validated
| null | null |
null |
{
"abstract": " We establish a new connection between value and policy based reinforcement\nlearning (RL) based on a relationship between softmax temporal value\nconsistency and policy optimality under entropy regularization. Specifically,\nwe show that softmax consistent action values correspond to optimal entropy\nregularized policy probabilities along any action sequence, regardless of\nprovenance. From this observation, we develop a new RL algorithm, Path\nConsistency Learning (PCL), that minimizes a notion of soft consistency error\nalong multi-step action sequences extracted from both on- and off-policy\ntraces. We examine the behavior of PCL in different scenarios and show that PCL\ncan be interpreted as generalizing both actor-critic and Q-learning algorithms.\nWe subsequently deepen the relationship by showing how a single model can be\nused to represent both a policy and the corresponding softmax state values,\neliminating the need for a separate critic. The experimental evaluation\ndemonstrates that PCL significantly outperforms strong actor-critic and\nQ-learning baselines across several benchmarks.\n",
"title": "Bridging the Gap Between Value and Policy Based Reinforcement Learning"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6149
| null |
Validated
| null | null |
null |
{
"abstract": " Two-dimensional (2-D) materials are of tremendous interest to integrated\nphotonics given their singular optical characteristics spanning light emission,\nmodulation, saturable absorption, and nonlinear optics. To harness their\noptical properties, these atomically thin materials are usually attached onto\nprefabricated devices via a transfer process. In this paper, we present a new\nroute for 2-D material integration with planar photonics. Central to this\napproach is the use of chalcogenide glass, a multifunctional material which can\nbe directly deposited and patterned on a wide variety of 2-D materials and can\nsimultaneously function as the light guiding medium, a gate dielectric, and a\npassivation layer for 2-D materials. Besides claiming improved fabrication\nyield and throughput compared to the traditional transfer process, our\ntechnique also enables unconventional multilayer device geometries optimally\ndesigned for enhancing light-matter interactions in the 2-D layers.\nCapitalizing on this facile integration method, we demonstrate a series of\nhigh-performance glass-on-graphene devices including ultra-broadband on-chip\npolarizers, energy-efficient thermo-optic switches, as well as graphene-based\nmid-infrared (mid-IR) waveguide-integrated photodetectors and modulators.\n",
"title": "Chalcogenide Glass-on-Graphene Photonics"
}
| null | null | null | null | true | null |
6150
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we investigate the use of adversarial domain adaptation for\naddressing the problem of language mismatch between speaker recognition\ncorpora. In the context of speaker verification, adversarial domain adaptation\nmethods aim at minimizing certain divergences between the distribution that the\nutterance-level features follow (i.e. speaker embeddings) when drawn from\nsource and target domains (i.e. languages), while preserving their capacity in\nrecognizing speakers. Neural architectures for extracting utterance-level\nrepresentations enable us to apply adversarial adaptation methods in an\nend-to-end fashion and train the network jointly with the standard\ncross-entropy loss. We examine several configurations, such as the use of\n(pseudo-)labels on the target domain as well as domain labels in the feature\nextractor, and we demonstrate the effectiveness of our method on the\nchallenging NIST SRE16 and SRE18 benchmarks.\n",
"title": "Speaker verification using end-to-end adversarial language adaptation"
}
| null | null | null | null | true | null |
6151
| null |
Default
| null | null |
null |
{
"abstract": " We study the elliptic curve $E_a: (ax+1)y^2+(ax+1)(x-1)y+x^2-x=0$, which we\ncall the geometric normal form of an elliptic curve. We show that any elliptic\ncurve whose $j$-invariant is real is isomorphic to a curve $E_a$ in geometric\nnormal form, and show that for $a \\notin \\{0, -1, -9\\}$, the points on $E_a$,\nminus a set of $6$ points, can be characterized in terms of the cevian geometry\nof a triangle.\n",
"title": "Real elliptic curves and cevian geometry"
}
| null | null | null | null | true | null |
6152
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we introduce the notion of an omni $n$-Lie algebra and show\nthat they are linearization of higher analogues of standard Courant algebroids.\nWe also introduce the notion of a nonabelian omni $n$-Lie algebra and show that\nthey are linearization of higher analogues of Courant algebroids associated to\nNambu-Poisson manifolds.\n",
"title": "Omni $n$-Lie algebras and linearization of higher analogues of Courant algebroids"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6153
| null |
Validated
| null | null |
null |
{
"abstract": " In this study an Artificial Neural Network was trained to classify musical\ninstruments, using audio samples transformed to the frequency domain. Different\nfeatures of the sound, in both time and frequency domain, were analyzed and\ncompared in relation to how much information that could be derived from that\nlimited data. The study concluded that in comparison with the base experiment,\nthat had an accuracy of 93.5%, using the attack only resulted in 80.2% and the\ninitial 100 Hz in 64.2%.\n",
"title": "Musical Instrument Recognition Using Their Distinctive Characteristics in Artificial Neural Networks"
}
| null | null | null | null | true | null |
6154
| null |
Default
| null | null |
null |
{
"abstract": " Deep convolutional neural networks (CNNs) have demonstrated impressive\nperformance on visual object classification tasks. In addition, it is a useful\nmodel for predication of neuronal responses recorded in visual system. However,\nthere is still no clear understanding of what CNNs learn in terms of visual\nneuronal circuits. Visualizing CNN's features to obtain possible connections to\nneuronscience underpinnings is not easy due to highly complex circuits from the\nretina to higher visual cortex. Here we address this issue by focusing on\nsingle retinal ganglion cells with a simple model and electrophysiological\nrecordings from salamanders. By training CNNs with white noise images to\npredicate neural responses, we found that convolutional filters learned in the\nend are resembling to biological components of the retinal circuit. Features\nrepresented by these filters tile the space of conventional receptive field of\nretinal ganglion cells. These results suggest that CNN could be used to reveal\nstructure components of neuronal circuits.\n",
"title": "Revealing structure components of the retina by deep learning networks"
}
| null | null | null | null | true | null |
6155
| null |
Default
| null | null |
null |
{
"abstract": " By a generalized Yangian we mean a Yangian-like algebra of one of two\nclasses. One of these classes consists of the so-called braided Yangians,\nintroduced in our previous paper. The braided Yangians are in a sense similar\nto the reflection equation algebra. The generalized Yangians of second class,\ncalled the Yangians of RTT type, are defined by the same formulae as the usual\nYangians are but with other quantum $R$-matrices. If such an $R$-matrix is the\nsimplest trigonometrical $R$-matrix, the corresponding Yangian of RTT type is\nthe so-called q-Yangian. We claim that each generalized Yangian is a\ndeformation of the commutative algebra ${\\rm Sym}(gl(m)[t^{-1}])$ provided that\nthe corresponding $R$-matrix is a deformation of the flip. Also, we exhibit the\ncorresponding Poisson brackets.\n",
"title": "Generalized Yangians and their Poisson counterparts"
}
| null | null | null | null | true | null |
6156
| null |
Default
| null | null |
null |
{
"abstract": " We obtain asymptotics of large Hankel determinants whose weight depends on a\none-cut regular potential and any number of Fisher-Hartwig singularities. This\ngeneralises two results: 1) a result of Berestycki, Webb and Wong [5] for\nroot-type singularities, and 2) a result of Its and Krasovsky [37] for a\nGaussian weight with a single jump-type singularity. We show that when we apply\na piecewise constant thinning on the eigenvalues of a random Hermitian matrix\ndrawn from a one-cut regular ensemble, the gap probability in the thinned\nspectrum, as well as correlations of the characteristic polynomial of the\nassociated conditional point process, can be expressed in terms of these\ndeterminants.\n",
"title": "Asymptotics of Hankel determinants with a one-cut regular potential and Fisher-Hartwig singularities"
}
| null | null | null | null | true | null |
6157
| null |
Default
| null | null |
null |
{
"abstract": " Galaxy intrinsic alignments (IA) are a critical uncertainty for current and\nfuture weak lensing measurements. We describe a perturbative expansion of IA,\nanalogous to the treatment of galaxy biasing. From an astrophysical\nperspective, this model includes the expected large-scale alignment mechanisms\nfor galaxies that are pressure-supported (tidal alignment) and\nrotation-supported (tidal torquing) as well as the cross-correlation between\nthe two. Alternatively, this expansion can be viewed as an effective model\ncapturing all relevant effects up to the given order. We include terms up to\nsecond order in the density and tidal fields and calculate the resulting IA\ncontributions to two-point statistics at one-loop order. For fiducial\namplitudes of the IA parameters, we find the quadratic alignment and\nlinear-quadratic cross terms can contribute order-unity corrections to the\ntotal intrinsic alignment signal at $k\\sim0.1\\,h^{-1}{\\rm Mpc}$, depending on\nthe source redshift distribution. These contributions can lead to significant\nbiases on inferred cosmological parameters in Stage IV photometric weak lensing\nsurveys. We perform forecasts for an LSST-like survey, finding that use of the\nstandard \"NLA\" model for intrinsic alignments cannot remove these large\nparameter biases, even when allowing for a more general redshift dependence.\nThe model presented here will allow for more accurate and flexible IA treatment\nin weak lensing and combined probes analyses, and an implementation is made\navailable as part of the public FAST-PT code. The model also provides a more\nadvanced framework for understanding the underlying IA processes and their\nrelationship to fundamental physics.\n",
"title": "Beyond linear galaxy alignments"
}
| null | null | null | null | true | null |
6158
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we introduce a new mathematical model for the active\ncontraction of cardiac muscle, featuring different thermo-electric and\nnonlinear conductivity properties. The passive hyperelastic response of the\ntissue is described by an orthotropic exponential model, whereas the ionic\nactivity dictates active contraction incorporated through the concept of\northotropic active strain. We use a fully incompressible formulation, and the\ngenerated strain modifies directly the conductivity mechanisms in the medium\nthrough the pull-back transformation. We also investigate the influence of\nthermo-electric effects in the onset of multiphysics emergent spatiotemporal\ndynamics, using nonlinear diffusion. It turns out that these ingredients have a\nkey role in reproducing pathological chaotic dynamics such as ventricular\nfibrillation during inflammatory events, for instance. The specific structure\nof the governing equations suggests to cast the problem in mixed-primal form\nand we write it in terms of Kirchhoff stress, displacements, solid pressure,\nelectric potential, activation generation, and ionic variables. We also propose\na new mixed-primal finite element method for its numerical approximation, and\nwe use it to explore the properties of the model and to assess the importance\nof coupling terms, by means of a few computational experiments in 3D.\n",
"title": "Modelling thermo-electro-mechanical effects in orthotropic cardiac tissue"
}
| null | null | null | null | true | null |
6159
| null |
Default
| null | null |
null |
{
"abstract": " Fixing bugs is an important phase in software development and maintenance. In\npractice, the process of bug fixing may conflict with the release schedule.\nSuch confliction leads to a trade-off between software quality and release\nschedule, which is known as the technical debt metaphor. In this article, we\npropose the concept of debt-prone bugs to model the technical debt in software\nmaintenance. We identify three types of debt-prone bugs, namely tag bugs,\nreopened bugs, and duplicate bugs. A case study on Mozilla is conducted to\nexamine the impact of debt-prone bugs in software products. We investigate the\ncorrelation between debt-prone bugs and the product quality. For a product\nunder development, we build prediction models based on historical products to\npredict the time cost of fixing bugs. The result shows that identifying\ndebt-prone bugs can assist in monitoring and improving software quality.\n",
"title": "Debt-Prone Bugs: Technical Debt in Software Maintenance"
}
| null | null | null | null | true | null |
6160
| null |
Default
| null | null |
null |
{
"abstract": " We give a simple recursion which computes the triply graded Khovanov-Rozansky\nhomology of several infinite families of knots and links, including the\n$(n,nm\\pm 1)$ and $(n,nm)$ torus links for $n,m\\geq 1$. We interpret our\nresults in terms of Catalan combinatorics, proving a conjecture of Gorsky's.\nOur computations agree with predictions coming from Hilbert schemes and\nrational DAHA, which also proves the Gorsky-Oblomkov-Rasmussen-Shende\nconjectures in these cases. Additionally, our results suggest a topological\ninterpretation of the symmetric functions which appear in the context of the\n$m$-shuffle conjecture of Haglund-Haiman-Loehr-Remmel-Ulyanov.\n",
"title": "Khovanov-Rozansky homology and higher Catalan sequences"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6161
| null |
Validated
| null | null |
null |
{
"abstract": " We develop an extended multifractal analysis based on the Legendre-Fenchel\ntransform rather than the routinely used Legendre transform. We apply this\nanalysis to studying time series consisting of inter-event times. As a result,\nwe discern the non-monotonic behavior of the generalized Hurst exponent - the\nfundamental exponent studied by us - and hence a multi-branched left-sided\nspectrum of dimensions. This kind of multifractality is a direct result of the\nnon-monotonic behavior of the generalized Hurst exponent and is not caused by\nnon-analytic behavior as has been previously suggested. We examine the main\nthermodynamic consequences of the existence of this type of multifractality\nrelated to the thermal stable, metastable, and unstable phases within a\nhierarchy of fluctuations, and also to the first and second order phase\ntransitions between them.\n",
"title": "New face of multifractality: Multi-branched left-sidedness and phase transitions in multifractality of interevent times"
}
| null | null |
[
"Quantitative Finance"
] | null | true | null |
6162
| null |
Validated
| null | null |
null |
{
"abstract": " By building up on the recent theory that established the connection between\nimplicit generative modeling and optimal transport, in this study, we propose a\nnovel parameter-free algorithm for learning the underlying distributions of\ncomplicated datasets and sampling from them. The proposed algorithm is based on\na functional optimization problem, which aims at finding a measure that is\nclose to the data distribution as much as possible and also expressive enough\nfor generative modeling purposes. We formulate the problem as a gradient flow\nin the space of probability measures. The connections between gradient flows\nand stochastic differential equations let us develop a computationally\nefficient algorithm for solving the optimization problem, where the resulting\nalgorithm resembles the recent dynamics-based Markov Chain Monte Carlo\nalgorithms. We provide formal theoretical analysis where we prove finite-time\nerror guarantees for the proposed algorithm. Our experimental results support\nour theory and shows that our algorithm is able to capture the structure of\nchallenging distributions.\n",
"title": "Sliced-Wasserstein Flows: Nonparametric Generative Modeling via Optimal Transport and Diffusions"
}
| null | null | null | null | true | null |
6163
| null |
Default
| null | null |
null |
{
"abstract": " Iterative Hard Thresholding (IHT) is a class of projected gradient descent\nmethods for optimizing sparsity-constrained minimization models, with the best\nknown efficiency and scalability in practice. As far as we know, the existing\nIHT-style methods are designed for sparse minimization in primal form. It\nremains open to explore duality theory and algorithms in such a non-convex and\nNP-hard problem setting. In this paper, we bridge this gap by establishing a\nduality theory for sparsity-constrained minimization with $\\ell_2$-regularized\nloss function and proposing an IHT-style algorithm for dual maximization. Our\nsparse duality theory provides a set of sufficient and necessary conditions\nunder which the original NP-hard/non-convex problem can be equivalently solved\nin a dual formulation. The proposed dual IHT algorithm is a super-gradient\nmethod for maximizing the non-smooth dual objective. An interesting finding is\nthat the sparse recovery performance of dual IHT is invariant to the Restricted\nIsometry Property (RIP), which is required by virtually all the existing primal\nIHT algorithms without sparsity relaxation. Moreover, a stochastic variant of\ndual IHT is proposed for large-scale stochastic optimization. Numerical results\ndemonstrate the superiority of dual IHT algorithms to the state-of-the-art\nprimal IHT-style algorithms in model estimation accuracy and computational\nefficiency.\n",
"title": "Dual Iterative Hard Thresholding: From Non-convex Sparse Minimization to Non-smooth Concave Maximization"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6164
| null |
Validated
| null | null |
null |
{
"abstract": " Binary neural networks (BNN) have been studied extensively since they run\ndramatically faster at lower memory and power consumption than floating-point\nnetworks, thanks to the efficiency of bit operations. However, contemporary\nBNNs whose weights and activations are both single bits suffer from severe\naccuracy degradation. To understand why, we investigate the representation\nability, speed and bias/variance of BNNs through extensive experiments. We\nconclude that the error of BNNs is predominantly caused by the intrinsic\ninstability (training time) and non-robustness (train & test time). Inspired by\nthis investigation, we propose the Binary Ensemble Neural Network (BENN) which\nleverages ensemble methods to improve the performance of BNNs with limited\nefficiency cost. While ensemble techniques have been broadly believed to be\nonly marginally helpful for strong classifiers such as deep neural networks,\nour analyses and experiments show that they are naturally a perfect fit to\nboost BNNs. We find that our BENN, which is faster and much more robust than\nstate-of-the-art binary networks, can even surpass the accuracy of the\nfull-precision floating number network with the same architecture.\n",
"title": "Binary Ensemble Neural Network: More Bits per Network or More Networks per Bit?"
}
| null | null |
[
"Statistics"
] | null | true | null |
6165
| null |
Validated
| null | null |
null |
{
"abstract": " We define the generalized connected sum for generic closed plane curves,\ngeneralizing the strange sum defined by Arnold, and completely describe how the\nArnold invariants $J^{\\pm}$ and $\\mathit{St}$ behave under the generalized\nconnected sums.\n",
"title": "Generalized connected sum formula for the Arnold invariants of generic plane curves"
}
| null | null | null | null | true | null |
6166
| null |
Default
| null | null |
null |
{
"abstract": " The emissivity of common materials remains constant with temperature\nvariations, and cannot drastically change. However, it is possible to design\nits entire behaviour as a function of temperature, and to significantly modify\nthe thermal emissivity of a surface through the combination of different\nmaterials and patterns. Here, we show that smart patterned surfaces consisting\nof smaller structures (motifs) may be designed to respond uniquely through\ncombinatorial design strategies by transforming themselves from 2D to 3D\ncomplex structures with a two-way shape memory effect. The smart surfaces can\npassively manipulate thermal radiation without-the use of controllers and power\nsupplies-because their modus operandi has already been programmed and\nintegrated into their intrinsic characteristics; the environment provides the\nenergy required for their activation. Each motif emits thermal radiation in a\ncertain manner, as it changes its geometry; however, the spatial distribution\nof these motifs causes them to interact with each other. Therefore, their\ncombination and interaction determine the global behaviour of the surfaces,\nthus enabling their a priori design. The emissivity behaviour is not random; it\nis determined by two fundamental parameters, namely the combination of\norientations in which the motifs open (n-fold rotational symmetry (rn)) and the\ncombination of materials (colours) on the motifs; these generate functions\nwhich fully determine the dependency of the emissivity on the temperature.\n",
"title": "Smart patterned surfaces with programmable thermal emissivity and their design through combinatorial strategies"
}
| null | null | null | null | true | null |
6167
| null |
Default
| null | null |
null |
{
"abstract": " Satellite radar altimetry is one of the most powerful techniques for\nmeasuring sea surface height variations, with applications ranging from\noperational oceanography to climate research. Over open oceans, altimeter\nreturn waveforms generally correspond to the Brown model, and by inversion,\nestimated shape parameters provide mean surface height and wind speed. However,\nin coastal areas or over inland waters, the waveform shape is often distorted\nby land influence, resulting in peaks or fast decaying trailing edges. As a\nresult, derived sea surface heights are then less accurate and waveforms need\nto be reprocessed by sophisticated algorithms. To this end, this work suggests\na novel Spatio-Temporal Altimetry Retracking (STAR) technique. We show that\nSTAR enables the derivation of sea surface heights over the open ocean as well\nas over coastal regions of at least the same quality as compared to existing\nretracking methods, but for a larger number of cycles and thus retaining more\nuseful data. Novel elements of our method are (a) integrating information from\nspatially and temporally neighboring waveforms through a conditional random\nfield approach, (b) sub-waveform detection, where relevant sub-waveforms are\nseparated from corrupted or non-relevant parts through a sparse representation\napproach, and (c) identifying the final best set of sea surfaces heights from\nmultiple likely heights using Dijkstra's algorithm. We apply STAR to data from\nthe Jason-1, Jason-2 and Envisat missions for study sites in the Gulf of\nTrieste, Italy and in the coastal region of the Ganges-Brahmaputra-Meghna\nestuary, Bangladesh. We compare to several established and recent retracking\nmethods, as well as to tide gauge data. Our experiments suggest that the\nobtained sea surface heights are significantly less affected by outliers when\ncompared to results obtained by other approaches.\n",
"title": "STAR: Spatio-Temporal Altimeter Waveform Retracking using Sparse Representation and Conditional Random Fields"
}
| null | null | null | null | true | null |
6168
| null |
Default
| null | null |
null |
{
"abstract": " We investigate the adiabatic magnetization process of the one-dimensional\n$J-Q_{2}$ model with XXZ anisotropy $g$ in an external magnetic field $h$ by\nusing density matrix renormalization group (DMRG) method. According to the\ncharacteristic of the magnetization curves, we draw a magnetization phase\ndiagram consisting of four phases. For a fixed nonzero pair coupling $Q$, i)\nwhen $g<-1$, the ground state is always ferromagnetic in spite of $h$; ii) when\n$g>-1$ but still small, the whole magnetization curve is continuous and smooth;\niii) if further increasing $g$, there is a macroscopic magnetization jump from\npartially- to fully-polarized state; iv) for a sufficiently large $g$, the\nmagnetization jump is from non- to fully-polarized state. By examining the\nenergy per magnon and the correlation function, we find that the origin of the\nmagnetization jump is the condensation of magnons and the formation of magnetic\ndomains. We also demonstrate that while the experienced states are\nHeisenberg-like without long-range order, all the \\textit{jumped-over} states\nhave antiferromagnetic or Néel long-range orders, or their mixing.\n",
"title": "Magnetization jump in one dimensional $J-Q_{2}$ model with anisotropic exchange"
}
| null | null | null | null | true | null |
6169
| null |
Default
| null | null |
null |
{
"abstract": " Cyber-physical software continually interacts with its physical environment\nfor adaptation in order to deliver smart services. However, the interactions\ncan be subject to various errors when the software's assumption on its\nenvironment no longer holds, thus leading to unexpected misbehavior or even\nfailure. To address this problem, one promising way is to conduct runtime\nmonitoring of invariants, so as to prevent cyber-physical software from\nentering such errors (a.k.a. abnormal states). To effectively detect abnormal\nstates, we in this article present an approach, named Context-based\nMulti-Invariant Detection (CoMID), which consists of two techniques:\ncontext-based trace grouping and multi-invariant detection. The former infers\ncontexts to distinguish different effective scopes for CoMID's derived\ninvariants, and the latter conducts ensemble evaluation of multiple invariants\nto detect abnormal states. We experimentally evaluate CoMID on real-world\ncyber-physical software. The results show that CoMID achieves a 5.7-28.2%\nhigher true-positive rate and a 6.8-37.6% lower false-positive rate in\ndetecting abnormal states, as compared with state-of-the-art approaches (i.e.,\nDaikon and ZoomIn). When deployed in field tests, CoMID's runtime monitoring\nimproves the success rate of cyber-physical software in its task executions by\n15.3-31.7%.\n",
"title": "CoMID: Context-based Multi-Invariant Detection for Monitoring Cyber-Physical Software"
}
| null | null | null | null | true | null |
6170
| null |
Default
| null | null |
null |
{
"abstract": " We consider the asymptotic distribution of a cell in a 2 x ... x 2\ncontingency table as the fixed marginal totals tend to infinity. The asymptotic\norder of the cell variance is derived and a useful diagnostic is given for\ndetermining whether the cell has a Poisson limit or a Gaussian limit. There are\nthree forms of Poisson convergence. The exact form is shown to be determined by\nthe growth rates of the two smallest marginal totals. The results are\ngeneralized to contingency tables with arbitrary sizes and are further\ncomplemented with concrete examples.\n",
"title": "Asymptotics of multivariate contingency tables with fixed marginals"
}
| null | null | null | null | true | null |
6171
| null |
Default
| null | null |
null |
{
"abstract": " This paper explores 1-dimensional topological quantum field theories. We\nseparately deal with strict and strong 1-dimensional topological quantum field\ntheories. The strict one is regarded as a symmetric monoidal functor between\nthe category of 1-cobordisms and the category of matrices, and the strong one\nis a symmetric monoidal functor between the category of 1-cobordisms and the\ncategory of finite dimensional vector spaces. It has been proved that both\nstrict and strong 1-dimensional topological quantum field theories are\nfaithful.\n",
"title": "On the Faithfulness of 1-dimensional Topological Quantum Field Theories"
}
| null | null | null | null | true | null |
6172
| null |
Default
| null | null |
null |
{
"abstract": " We expand the cross section of the geodesic flow in the tangent bundle of the\nmodular surface given by Series to produce another section whose return map\nunder the geodesic flow is a double cover of the natural extension of the Farey\nmap. We use this cross section to extend the correspondence between the closed\ngeodesics on the modular surface and the periodic points of the Gauss map to\ninclude the periodic points of the Farey map. Then, analogous to the work of\nPollicott, we prove an equidistribution result for the periodic points of the\nFarey map when they are ordered according to the length of their corresponding\nclosed geodesics.\n",
"title": "Distribution of the periodic points of the Farey map"
}
| null | null | null | null | true | null |
6173
| null |
Default
| null | null |
null |
{
"abstract": " We study the time evolution of a thin liquid film coating the outer surface\nof a sphere in the presence of gravity, surface tension and thermal gradients.\nWe derive the fourth-order nonlinear partial differential equation that models\nthe thin film dynamics, including Marangoni terms arising from the dependence\nof surface tension on temperature. We consider two different imposed\ntemperature distributions with axial or radial thermal gradients. We analyze\nthe stability of a uniform coating under small perturbations and carry out\nnumerical simulations in COMSOL for a range of parameter values. In the case of\nan axial temperature gradient, we find steady states with either uniform film\nthickness, or with the fluid accumulating at the bottom or near the top of the\nsphere, depending on the total volume of liquid in the film, dictating whether\ngravity or Marangoni effects dominate. In the case of a radial temperature\ngradient, a stability analysis reveals the most unstable non-axisymmetric modes\non an initially uniform coating film.\n",
"title": "Marangoni effects on a thin liquid film coating a sphere with axial or radial thermal gradients"
}
| null | null | null | null | true | null |
6174
| null |
Default
| null | null |
null |
{
"abstract": " Machine learning algorithms are sensitive to so-called adversarial\nperturbations. This is reminiscent of cellular decision-making where antagonist\nligands may prevent correct signaling, like during the early immune response.\nWe draw a formal analogy between neural networks used in machine learning and\nthe general class of adaptive proofreading networks. We then apply simple\nadversarial strategies from machine learning to models of ligand\ndiscrimination. We show how kinetic proofreading leads to \"boundary tilting\"\nand identify three types of perturbation (adversarial, non adversarial and\nambiguous). We then use a gradient-descent approach to compare different\nadaptive proofreading models, and we reveal the existence of two qualitatively\ndifferent regimes characterized by the presence or absence of a critical point.\nThese regimes are reminiscent of the \"feature-to-prototype\" transition\nidentified in machine learning, corresponding to two strategies in ligand\nantagonism (broad vs. specialized). Overall, our work connects evolved cellular\ndecision-making to classification in machine learning, showing that behaviours\nclose to the decision boundary can be understood through the same mechanisms.\n",
"title": "Fooling the classifier: Ligand antagonism and adversarial examples"
}
| null | null | null | null | true | null |
6175
| null |
Default
| null | null |
null |
{
"abstract": " A simplified approach is proposed to investigate the continuous-time and\ndiscrete-time complementary sensitivity Bode integrals (CSBIs) in this note.\nFor continuous-time feedback systems with unbounded frequency domain, the CSBI\nweighted by $1/\\omega^2$ is considered, where this simplified method reveals a\nmore explicit relationship between the value of CSBI and the structure of the\nopen-loop transfer function. With a minor modification of this method, the CSBI\nof discrete-time system is derived, and illustrative examples are provided.\nCompared with the existing results on CSBI, neither Cauchy integral theorem nor\nPoisson integral formula are used throughout the analysis, and the analytic\nconstraint on the integrand is removed.\n",
"title": "A Simplified Approach to Analyze Complementary Sensitivity Trade-offs in Continuous-Time and Discrete-Time Systems"
}
| null | null | null | null | true | null |
6176
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we propose a family of graph partition similarity measures\nthat take the topology of the graph into account. These graph-aware measures\nare alternatives to using set partition similarity measures that are not\nspecifically designed for graph partitions. The two types of measures,\ngraph-aware and set partition measures, are shown to have opposite behaviors\nwith respect to resolution issues and provide complementary information\nnecessary to assess that two graph partitions are similar.\n",
"title": "Comparing Graph Clusterings: Set partition measures vs. Graph-aware measures"
}
| null | null | null | null | true | null |
6177
| null |
Default
| null | null |
null |
{
"abstract": " Heterogeneous wireless networks with small-cell deployments in licensed and\nunlicensed spectrum bands are a promising approach for expanding wireless\nconnectivity and service. As a result, wireless service providers (SPs) are\nadding small-cells to augment their existing macro-cell deployments. This added\nflexibility complicates network management, in particular, service pricing and\nspectrum allocations across macro- and small-cells. Further, these decisions\ndepend on the degree of competition among SPs. Restrictions on shared spectrum\naccess imposed by regulators, such as low power constraints that lead to\nsmall-cell deployments, along with the investment cost needed to add small\ncells to an existing network, also impact strategic decisions and market\nefficiency. If the revenue generated by small-cells does not cover the\ninvestment cost, then there will be no deployment even if it increases social\nwelfare. We study the implications of such spectrum constraints and investment\ncosts on resource allocation and pricing decisions by competitive SPs, along\nwith the associated social welfare. Our results show that while the optimal\nresource allocation taking constraints and investment into account can be\nuniquely determined, adding those features with strategic SPs can have a\nsubstantial effect on the equilibrium market structure.\n",
"title": "Competitive Resource Allocation in HetNets: the Impact of Small-cell Spectrum Constraints and Investment Costs"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6178
| null |
Validated
| null | null |
null |
{
"abstract": " No methods currently exist for making arbitrary neural networks fair. In this\nwork we introduce GRAD, a new and simplified method to producing fair neural\nnetworks that can be used for auto-encoding fair representations or directly\nwith predictive networks. It is easy to implement and add to existing\narchitectures, has only one (insensitive) hyper-parameter, and provides\nimproved individual and group fairness. We use the flexibility of GRAD to\ndemonstrate multi-attribute protection.\n",
"title": "Gradient Reversal Against Discrimination"
}
| null | null | null | null | true | null |
6179
| null |
Default
| null | null |
null |
{
"abstract": " We give a criterion for the existence of non-commutative crepant resolutions\n(NCCR's) for certain toric singularities. In particular we recover Broomhead's\nresult that a 3-dimensional toric Gorenstein singularity has a NCCR. Our result\nalso yields the existence of a NCCR for a 4-dimensional toric Gorenstein\nsingularity which is known to have no toric NCCR.\n",
"title": "Non-commutative crepant resolutions for some toric singularities I"
}
| null | null | null | null | true | null |
6180
| null |
Default
| null | null |
null |
{
"abstract": " We observe many-body pairing in a two-dimensional gas of ultracold fermionic\natoms at temperatures far above the critical temperature for superfluidity. For\nthis, we use spatially resolved radio-frequency spectroscopy to measure pairing\nenergies spanning a wide range of temperatures and interaction strengths. In\nthe strongly interacting regime where the scattering length between fermions is\non the same order as the inter-particle spacing, the pairing energy in the\nnormal phase significantly exceeds the intrinsic two-body binding energy of the\nsystem and shows a clear dependence on local density. This implies that pairing\nin this regime is driven by many-body correlations, rather than two-body\nphysics. We find this effect to persist at temperatures close to the Fermi\ntemperature which demonstrates that pairing correlations in strongly\ninteracting two-dimensional fermionic systems are remarkably robust against\nthermal fluctuations.\n",
"title": "High temperature pairing in a strongly interacting two-dimensional Fermi gas"
}
| null | null | null | null | true | null |
6181
| null |
Default
| null | null |
null |
{
"abstract": " Complex mathematical models of interaction networks are routinely used for\nprediction in systems biology. However, it is difficult to reconcile network\ncomplexities with a formal understanding of their behavior. Here, we propose a\nsimple procedure (called $\\bar \\varphi$) to reduce biological models to\nfunctional submodules, using statistical mechanics of complex systems combined\nwith a fitness-based approach inspired by $\\textit{in silico}$ evolution. $\\bar\n\\varphi$ works by putting parameters or combination of parameters to some\nasymptotic limit, while keeping (or slightly improving) the model performance,\nand requires parameter symmetry breaking for more complex models. We illustrate\n$\\bar \\varphi$ on biochemical adaptation and on different models of immune\nrecognition by T cells. An intractable model of immune recognition with close\nto a hundred individual transition rates is reduced to a simple two-parameter\nmodel. $\\bar \\varphi$ extracts three different mechanisms for early immune\nrecognition, and automatically discovers similar functional modules in\ndifferent models of the same process, allowing for model classification and\ncomparison. Our procedure can be applied to biological networks based on rate\nequations using a fitness function that quantifies phenotypic performance.\n",
"title": "Untangling the hairball: fitness based asymptotic reduction of biological networks"
}
| null | null | null | null | true | null |
6182
| null |
Default
| null | null |
null |
{
"abstract": " An immense class of physical counterexamples to the four dimensional strong\ncosmic censor conjecture---in its usual broad formulation---is exhibited. More\nprecisely, out of any closed and simply connected 4-manifold an open Ricci-flat\nLorentzian 4-manifold is constructed which is not globally hyperbolic and no\nperturbation of it, in any sense, can be globally hyperbolic. This very stable\nnon-global-hyperbolicity is the consequence of our open spaces having a\n\"creased end\" i.e., an end diffeomorphic to an exotic ${\\mathbb R}^4$. Open\nmanifolds having an end like this is a typical phenomenon in four dimensions.\nThe construction is based on a collection of results of Gompf and Taubes on\nexotic and self-dual spaces, respectively, as well as applying Penrose'\nnon-linear graviton construction (i.e., twistor theory) to solve the Riemannian\nEinstein's equation. These solutions then are converted into stably\nnon-globally-hyperbolic Lorentzian vacuum solutions. It follows that the\nplethora of vacuum solutions we found cannot be obtained via the initial value\nformulation of the Einstein's equation because they are \"too long\" in a certain\nsense (explained in the text). This different (i.e., not based on the initial\nvalue formulation but twistorial) technical background might partially explain\nwhy the existence of vacuum solutions of this kind have not been realized so\nfar in spite of the fact that, apparently, their superabundance compared to the\nwell-known globally hyperbolic vacuum solutions is overwhelming.\n",
"title": "Exotica and the status of the strong cosmic censor conjecture in four dimensions"
}
| null | null | null | null | true | null |
6183
| null |
Default
| null | null |
null |
{
"abstract": " Complex networks can be used to represent complex systems which originate in\nthe real world. Here we study a transformation of these complex networks into\nsimplicial complexes, where cliques represent the simplices of the complex. We\nextend the concept of node centrality to that of simplicial centrality and\nstudy several mathematical properties of degree, closeness, betweenness,\neigenvector, Katz, and subgraph centrality for simplicial complexes. We study\nthe degree distributions of these centralities at the different levels. We also\ncompare and describe the differences between the centralities at the different\nlevels. Using these centralities we study a method for detecting essential\nproteins in PPI networks of cells and explain the varying abilities of the\ncentrality measures at the different levels in identifying these essential\nproteins. The paper is written in a self-contained way, such that it can be\nused by practitioners of network theory as a basis for further developments.\n",
"title": "Centralities in Simplicial Complexes"
}
| null | null | null | null | true | null |
6184
| null |
Default
| null | null |
null |
{
"abstract": " For a uniform space (X, $\\mu$), we introduce a realcompactification of X by\nmeans of the family $U_{\\mu}(X)$ of all the real-valued uniformly continuous\nfunctions, in the same way that the known Samuel compactification is given by\n$U^{*}_{\\mu}(X)$ the set of all the bounded functions in $U_{\\mu}(X)$. We will\ncall it \"the Samuel realcompactification\" by several resemblances to the Samuel\ncompactification. In this note, we present different ways to construct such\nrealcompactification as well as we study the corresponding problem of knowing\nwhen a uniform space is Samuel realcompact, that is, it coincides with its\nSamuel realcompactification. At this respect we obtain as main result a theorem\nof Katětov-Shirota type, by means of a new property of completeness\nrecently introduced by the authors, called Bourbaki-completeness.\n",
"title": "The Samuel realcompactification"
}
| null | null | null | null | true | null |
6185
| null |
Default
| null | null |
null |
{
"abstract": " Acceleration and manipulation of ultrashort electron bunches are the basis\nbehind electron and X-ray devices used for ultrafast, atomic-scale imaging and\nspectroscopy. Using laser-generated THz drivers enables intrinsic\nsynchronization as well as dramatic gains in field strengths, field gradients\nand component compactness, leading to shorter electron bunches, higher\nspatio-temporal resolution and smaller infrastructures. We present a segmented\nTHz electron accelerator and manipulator (STEAM) with extended interaction\nlengths capable of performing multiple high-field operations on the energy and\nphase-space of ultrashort bunches with moderate charge. With this single\ndevice, powered by few-microjoule, single-cycle, 0.3 THz pulses, we demonstrate\nrecord THz-device acceleration of >30 keV, streaking with <10 fs resolution,\nfocusing with >2 kT/m strengths, compression to ~100 fs as well as real-time\nswitching between these modes of operation. The STEAM device demonstrates the\nfeasibility of future THz-based compact electron guns, accelerators, ultrafast\nelectron diffractometers and Free-Electron Lasers with transformative impact.\n",
"title": "Segmented Terahertz Electron Accelerator and Manipulator (STEAM)"
}
| null | null | null | null | true | null |
6186
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we derive the second order estimate to the $2$-nd Hessian type\nequation on a compact almost Hermitian manifold.\n",
"title": "The $2$-nd Hessian type equation on almost Hermitian manifolds"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6187
| null |
Validated
| null | null |
null |
{
"abstract": " The construction of a meaningful graph topology plays a crucial role in the\neffective representation, processing, analysis and visualization of structured\ndata. When a natural choice of the graph is not readily available from the data\nsets, it is thus desirable to infer or learn a graph topology from the data. In\nthis tutorial overview, we survey solutions to the problem of graph learning,\nincluding classical viewpoints from statistics and physics, and more recent\napproaches that adopt a graph signal processing (GSP) perspective. We further\nemphasize the conceptual similarities and differences between classical and\nGSP-based graph inference methods, and highlight the potential advantage of the\nlatter in a number of theoretical and practical scenarios. We conclude with\nseveral open issues and challenges that are keys to the design of future signal\nprocessing and machine learning algorithms for learning graphs from data.\n",
"title": "Learning graphs from data: A signal representation perspective"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6188
| null |
Validated
| null | null |
null |
{
"abstract": " We study the edge transport properties of $2d$ interacting Hall systems,\ndisplaying single-mode chiral edge currents. For this class of many-body\nlattice models, including for instance the interacting Haldane model, we prove\nthe quantization of the edge charge conductance and the bulk-edge\ncorrespondence. Instead, the edge Drude weight and the edge susceptibility are\ninteraction-dependent; nevertheless, they satisfy exact universal scaling\nrelations, in agreement with the chiral Luttinger liquid theory. Moreover,\ncharge and spin excitations differ in their velocities, giving rise to the\nspin-charge separation phenomenon. The analysis is based on exact\nrenormalization group methods, and on a combination of lattice and emergent\nWard identities. The invariance of the emergent chiral anomaly under the\nrenormalization group flow plays a crucial role in the proof.\n",
"title": "Universal edge transport in interacting Hall systems"
}
| null | null | null | null | true | null |
6189
| null |
Default
| null | null |
null |
{
"abstract": " We demonstrate experimentally that the long-range hydrodynamic interactions\nin an incompressible quasi 2D isotropic fluid result in an anisotropic viscous\ndrag acting on elongated particles. The anisotropy of the drag is increasing\nwith increasing ratio of the particle length to the hydrodynamic scale given by\nthe Saffman-Delbrück length. The micro-rheology data for translational and\nrotational drags collected over three orders of magnitude of the effective\nparticle length demonstrate the validity of the current theoretical approaches\nto the hydrodynamics in restricted geometry. The results also demonstrate\ncrossovers between the hydrodynamical regimes determined by the characteristic\nlength scales.\n",
"title": "Brownian dynamics of elongated particles in a quasi-2D isotropic liquid"
}
| null | null | null | null | true | null |
6190
| null |
Default
| null | null |
null |
{
"abstract": " Two-dimensional atomic arrays exhibit a number of intriguing quantum optical\nphenomena, including subradiance, nearly perfect reflection of radiation and\nlong-lived topological edge states. Studies of emission and scattering of\nphotons in such lattices require complete treatment of the radiation pattern\nfrom individual atoms, including long-range interactions. We describe a\nsystematic approach to perform the calculations of collective energy shifts and\ndecay rates in the presence of such long-range interactions for arbitrary\ntwo-dimensional atomic lattices. As applications of our method, we investigate\nthe topological properties of atomic lattices both in free-space and near\nplasmonic surfaces.\n",
"title": "Photonic Band Structure of Two-dimensional Atomic Lattices"
}
| null | null | null | null | true | null |
6191
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we applied the multifractal detrended fluctuation analysis to\nthe daily means of wind speed measured by 119 weather stations distributed over\nthe territory of Switzerland. The analysis was focused on the inner time\nfluctuations of wind speed, which could be more linked with the local\nconditions of the highly varying topography of Switzerland. Our findings point\nout to a persistent behaviour of all the measured wind speed series (indicated\nby a Hurst exponent significantly larger than 0.5), and to a high\nmultifractality degree indicating a relative dominance of the large\nfluctuations in the dynamics of wind speed, especially in the Swiss plateau,\nwhich is comprised between the Jura and Alp mountain ranges. The study\nrepresents a contribution to the understanding of the dynamical mechanisms of\nwind speed variability in mountainous regions.\n",
"title": "Multifractal analysis of the time series of daily means of wind speed in complex regions"
}
| null | null | null | null | true | null |
6192
| null |
Default
| null | null |
null |
{
"abstract": " We present a new deep meta reinforcement learner, which we call Deep Episodic\nValue Iteration (DEVI). DEVI uses a deep neural network to learn a similarity\nmetric for a non-parametric model-based reinforcement learning algorithm. Our\nmodel is trained end-to-end via back-propagation. Despite being trained using\nthe model-free Q-learning objective, we show that DEVI's model-based internal\nstructure provides `one-shot' transfer to changes in reward and transition\nstructure, even for tasks with very high-dimensional state spaces.\n",
"title": "Deep Episodic Value Iteration for Model-based Meta-Reinforcement Learning"
}
| null | null | null | null | true | null |
6193
| null |
Default
| null | null |
null |
{
"abstract": " Transiting super-Earths orbiting bright stars in short orbital periods are\ninteresting targets for the study of planetary atmospheres. While selecting\nsuper-Earths suitable for further characterization from the ground among a list\nof confirmed and validated exoplanets detected by K2, we found some suspicious\ncases that led to us re-assessing the nature of the detected transiting signal.\nWe did a photometric analysis of the K2 light curves and centroid motions of\nthe photometric barycenters. Our study shows that the validated planets K2-78b,\nK2-82b, and K2-92b are actually not planets but background eclipsing binaries.\nThe eclipsing binaries are inside the Kepler photometric aperture, but outside\nthe ground-based high resolution images used for validation. We advise extreme\ncare on the validation of candidate planets discovered by space missions. It is\nimportant that all the assumptions in the validation process are carefully\nchecked. An independent confirmation is mandatory in order to avoid wasting\nvaluable resources on further characterization of non-existent targets.\n",
"title": "Disproval of the validated planets K2-78b, K2-82b, and K2-92b"
}
| null | null | null | null | true | null |
6194
| null |
Default
| null | null |
null |
{
"abstract": " We consider two chains, each made of $N$ independent oscillators, immersed in\na common thermal bath and study the dynamics of their mutual quantum\ncorrelations in the thermodynamic, large-$N$ limit. We show that dissipation\nand noise due to the presence of the external environment are able to generate\ncollective quantum correlations between the two chains at the mesoscopic level.\nThe created collective quantum entanglement between the two many-body systems\nturns out to be rather robust, surviving for asymptotically long times even for\nnon vanishing bath temperatures.\n",
"title": "Long-lived mesoscopic entanglement between two damped infinite harmonic chains"
}
| null | null |
[
"Physics"
] | null | true | null |
6195
| null |
Validated
| null | null |
null |
{
"abstract": " A paper by Bruno Salvy and the author introduced measured multiseries and\ngave an algorithm to compute these for a large class of elementary functions,\nmodulo a zero-equivalence method for constants. This gave a theoretical\nbackground for the implementation that Salvy was developing at that time. The\nmain result of the present article is an algorithm to calculate measured\nmultiseries for integrals of functions of the form h*sin G, where h and G\nbelong to a Hardy field. The process can reiterated with the resulting algebra,\nand also applied to solutions of a second order differential equation of a\nparticular form.\n",
"title": "Measured Multiseries and Integration"
}
| null | null | null | null | true | null |
6196
| null |
Default
| null | null |
null |
{
"abstract": " This paper presents a fast and effective computer algebraic method for\nanalyzing and verifying non-linear integer arithmetic circuits using a novel\nalgebraic spectral model. It introduces a concept of algebraic spectrum, a\nnumerical form of polynomial expression; it uses the distribution of\ncoefficients of the monomials to determine the type of arithmetic function\nunder verification. In contrast to previous works, the proof of functional\ncorrectness is achieved by computing an algebraic spectrum combined with a\nlocal rewriting of word-level polynomials. The speedup is achieved by\npropagating coefficients through the circuit using And-Inverter Graph (AIG)\ndatastructure. The effectiveness of the method is demonstrated with experiments\nincluding standard and Booth multipliers, and other synthesized non-linear\narithmetic circuits up to 1024 bits containing over 12 million gates.\n",
"title": "Spectral Approach to Verifying Non-linear Arithmetic Circuits"
}
| null | null | null | null | true | null |
6197
| null |
Default
| null | null |
null |
{
"abstract": " Detecting weak seismic events from noisy sensors is a difficult perceptual\ntask. We formulate this task as Bayesian inference and propose a generative\nmodel of seismic events and signals across a network of spatially distributed\nstations. Our system, SIGVISA, is the first to directly model seismic\nwaveforms, allowing it to incorporate a rich representation of the physics\nunderlying the signal generation process. We use Gaussian processes over\nwavelet parameters to predict detailed waveform fluctuations based on\nhistorical events, while degrading smoothly to simple parametric envelopes in\nregions with no historical seismicity. Evaluating on data from the western US,\nwe recover three times as many events as previous work, and reduce mean\nlocation errors by a factor of four while greatly increasing sensitivity to\nlow-magnitude events.\n",
"title": "Signal-based Bayesian Seismic Monitoring"
}
| null | null |
[
"Computer Science",
"Physics"
] | null | true | null |
6198
| null |
Validated
| null | null |
null |
{
"abstract": " It is well-known that exploiting label correlations is important to\nmulti-label learning. Existing approaches either assume that the label\ncorrelations are global and shared by all instances; or that the label\ncorrelations are local and shared only by a data subset. In fact, in the\nreal-world applications, both cases may occur that some label correlations are\nglobally applicable and some are shared only in a local group of instances.\nMoreover, it is also a usual case that only partial labels are observed, which\nmakes the exploitation of the label correlations much more difficult. That is,\nit is hard to estimate the label correlations when many labels are absent. In\nthis paper, we propose a new multi-label approach GLOCAL dealing with both the\nfull-label and the missing-label cases, exploiting global and local label\ncorrelations simultaneously, through learning a latent label representation and\noptimizing label manifolds. The extensive experimental studies validate the\neffectiveness of our approach on both full-label and missing-label data.\n",
"title": "Multi-Label Learning with Global and Local Label Correlation"
}
| null | null | null | null | true | null |
6199
| null |
Default
| null | null |
null |
{
"abstract": " Complexity analysis becomes a common task in supervisory control. However,\nmany results of interest are spread across different topics. The aim of this\npaper is to bring several interesting results from complexity theory and to\nillustrate their relevance to supervisory control by proving new nontrivial\nresults concerning nonblockingness in modular supervisory control of discrete\nevent systems modeled by finite automata.\n",
"title": "Complexity of Verifying Nonblockingness in Modular Supervisory Control"
}
| null | null | null | null | true | null |
6200
| null |
Default
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.