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": " $k$-point crossover operators and their recombination sets are studied from\ndifferent perspectives. We show that transit functions of $k$-point crossover\ngenerate, for all $k>1$, the same convexity as the interval function of the\nunderlying graph. This settles in the negative an open problem by Mulder about\nwhether the geodesic convexity of a connected graph $G$ is uniquely determined\nby its interval function $I$. The conjecture of Gitchoff and Wagner that for\neach transit set $R_k(x,y)$ distinct from a hypercube there is a unique pair of\nparents from which it is generated is settled affirmatively. Along the way we\ncharacterize transit functions whose underlying graphs are Hamming graphs, and\nthose with underlying partial cube graphs. For general values of $k$ it is\nshown that the transit sets of $k$-point crossover operators are the subsets\nwith maximal Vapnik-Chervonenkis dimension. Moreover, the transit sets of\n$k$-point crossover on binary strings form topes of uniform oriented matroid of\nVC-dimension $k+1$. The Topological Representation Theorem for oriented\nmatroids therefore implies that $k$-point crossover operators can be\nrepresented by pseudosphere arrangements. This provides the tools necessary to\nstudy the special case $k=2$ in detail.\n",
"title": "Topological Representation of the Transit Sets of k-Point Crossover Operators"
}
| null | null | null | null | true | null |
13401
| null |
Default
| null | null |
null |
{
"abstract": " With the advances in robotic technology, research in human-robot\ncollaboration (HRC) has gained in importance. For robots to interact with\nhumans autonomously they need active decision making that takes human partners\ninto account. However, state-of-the-art research in HRC does often assume a\nleader-follower division, in which one agent leads the interaction. We believe\nthat this is caused by the lack of a reliable representation of the human and\nthe environment to allow autonomous decision making. This problem can be\novercome by an embodied approach to HRC which is inspired by psychological\nstudies of human-human interaction (HHI). In this survey, we review\nneuroscientific and psychological findings of the sensorimotor patterns that\ngovern HHI and view them in a robotics context. Additionally, we study the\nadvances made by the robotic community into the direction of embodied HRC. We\nfocus on the mechanisms that are required for active, physical human-robot\ncollaboration. Finally, we discuss the similarities and differences in the two\nfields of study which pinpoint directions of future research.\n",
"title": "Human-Robot Collaboration: From Psychology to Social Robotics"
}
| null | null | null | null | true | null |
13402
| null |
Default
| null | null |
null |
{
"abstract": " Recently, millimeter-wave (mmWave) communications have received great\nattention due to the availability of large spectrum resources. Nevertheless,\ntheir impact on TCP performance has been overlooked, which is observed that the\nsaid TCP performance collapse occurs owing to the significant difference in\nsignal quality between LOS and NLOS links. We propose a novel TCP design for\nmmWave communications, a mmWave performance enhancing proxy (mmPEP), enabling\nnot only to overcome TCP performance collapse but also exploit the properties\nof mmWave channels. The base station installs the TCP proxy to operate the two\nfunctionalities called Ack management and batch retransmission. Specifically,\nthe proxy sends the said early-Ack to the server not to decrease its sending\nrate even in the NLOS status. In addition, when a packet-loss is detected, the\nproxy retransmits not only lost packets but also the certain number of the\nfollowing packets expected to be lost too. It is verified by ns-3 simulation\nthat compared with benchmark, mmPEP enhances the end-to-end rate and packet\ndelivery ratio by maintaining high sending rate with decreasing the loss\nrecovery time.\n",
"title": "Enhancing TCP End-to-End Performance in Millimeter-Wave Communications"
}
| null | null | null | null | true | null |
13403
| null |
Default
| null | null |
null |
{
"abstract": " Quantum entanglement serves as a valuable resource for many important quantum\noperations. A pair of entangled qubits can be shared between two agents by\nfirst preparing a maximally entangled qubit pair at one agent, and then sending\none of the qubits to the other agent through a quantum channel. In this\nprocess, the deterioration of entanglement is inevitable since the noise\ninherent in the channel contaminates the qubit. To address this challenge,\nvarious quantum entanglement distillation (QED) algorithms have been developed.\nAmong them, recurrence algorithms have advantages in terms of implementability\nand robustness. However, the efficiency of recurrence QED algorithms has not\nbeen investigated thoroughly in the literature. This paper put forth two\nrecurrence QED algorithms that adapt to the quantum channel to tackle the\nefficiency issue. The proposed algorithms have guaranteed convergence for\nquantum channels with two Kraus operators, which include phase-damping and\namplitude-damping channels. Analytical results show that the convergence speed\nof these algorithms is improved from linear to quadratic and one of the\nalgorithms achieves the optimal speed. Numerical results confirm that the\nproposed algorithms significantly improve the efficiency of QED.\n",
"title": "Adaptive recurrence quantum entanglement distillation for two-Kraus-operator channels"
}
| null | null | null | null | true | null |
13404
| null |
Default
| null | null |
null |
{
"abstract": " We discuss the design and optimisation of two types of junctions between\nsurface-electrode radiofrequency ion-trap arrays that enable the integration of\nexperiments with sympathetically cooled molecular ions on a monolithic chip\ndevice. A detailed description of a multi-objective optimisation procedure\napplicable to an arbitrary planar junction is presented, and the results for a\ncross junction between four quadrupoles as well as a quadrupole-to-octupole\njunction are discussed. Based on these optimised functional elements, we\npropose a multi-functional ion-trap chip for experiments with translationally\ncold molecular ions at temperatures in the millikelvin range. This study opens\nthe door to extending complex chip-based trapping techniques to\nCoulomb-crystallised molecular ions with potential applications in mass\nspectrometry, spectroscopy, controlled chemistry and quantum technology.\n",
"title": "Optimised surface-electrode ion-trap junctions for experiments with cold molecular ions"
}
| null | null | null | null | true | null |
13405
| null |
Default
| null | null |
null |
{
"abstract": " We present a three-dimensional cubic lattice spin model, anisotropic in the\n$\\hat{z}$ direction, that exhibits fracton topological order. The latter is a\nnovel type of topological order characterized by the presence of immobile\npointlike excitations, named fractons, residing at the corners of an operator\nwith two-dimensional support. As other recent fracton models, ours exhibits a\nsubextensive ground state degeneracy: On an $L_x\\times L_y\\times L_z$\nthree-torus, it has a $2^{2L_z}$ topological degeneracy, and an additional\nnon-topological degeneracy equal to $2^{L_xL_y-2}$. The fractons can be\ncombined into composite excitations that move either in a straight line along\nthe $\\hat{z}$ direction, or freely in the $xy$ plane at a given height $z$.\nWhile our model draws inspiration from the toric code, we demonstrate that it\ncannot be adiabatically connected to a layered toric code construction.\nAdditionally, we investigate the effects of imposing open boundary conditions\non our system. We find zero energy modes on the surfaces perpendicular to\neither the $\\hat{x}$ or $\\hat{y}$ directions, and their absence on the surfaces\nnormal to $\\hat{z}$. This result can be explained using the properties of the\ntwo kinds of composite two-fracton mobile excitations.\n",
"title": "A simple anisotropic three-dimensional quantum spin liquid with fracton topological order"
}
| null | null | null | null | true | null |
13406
| null |
Default
| null | null |
null |
{
"abstract": " We advance the state of the art in polyphonic piano music transcription by\nusing a deep convolutional and recurrent neural network which is trained to\njointly predict onsets and frames. Our model predicts pitch onset events and\nthen uses those predictions to condition framewise pitch predictions. During\ninference, we restrict the predictions from the framewise detector by not\nallowing a new note to start unless the onset detector also agrees that an\nonset for that pitch is present in the frame. We focus on improving onsets and\noffsets together instead of either in isolation as we believe this correlates\nbetter with human musical perception. Our approach results in over a 100%\nrelative improvement in note F1 score (with offsets) on the MAPS dataset.\nFurthermore, we extend the model to predict relative velocities of normalized\naudio which results in more natural-sounding transcriptions.\n",
"title": "Onsets and Frames: Dual-Objective Piano Transcription"
}
| null | null | null | null | true | null |
13407
| null |
Default
| null | null |
null |
{
"abstract": " In a previous joint work of Xiao and the second author, the modified mean\ncurvature flow (MMCF) in hyperbolic space $\\mathbb{H}^{n+1}$: $$\\frac{\\partial\n\\mathbf{F}}{\\partial t} = (H-\\sigma)\\,\\vnu\\,,\\quad \\quad \\sigma\\in (-n,n)$$ was\nfirst introduced and the flow starting from an entire Lipschitz continuous\nradial graph with uniform local ball condition on the asymptotic boundary was\nshown to exist for all time and converge to a complete hypersurface of constant\nmean curvature with prescribed asymptotic boundary at infinity. In this paper,\nwe remove the uniform local ball condition on the asymptotic boundary of the\ninitial hypersurface, and prove that the MMCF starting from an entire locally\nLipschitz continuous radial graph exists and stays radially graphic for all\ntime.\n",
"title": "Modified mean curvature flow of entire locally Lipschitz radial graphs in hyperbolic space"
}
| null | null | null | null | true | null |
13408
| null |
Default
| null | null |
null |
{
"abstract": " We propose and demonstrate a method for calibrating atomic trajectories in a\nlarge-area dual-atom-interferometer gyroscope. The atom trajectories are\nmonitored by modulating and delaying the Raman transition, and they are\nprecisely calibrated by controlling the laser orientation and the bias magnetic\nfield. To improve the immunity to the gravity effect and the common phase\nnoise, the symmetry and the overlapping of two large-area atomic interference\nloops are optimized by calibrating the atomic trajectories and by aligning the\nRaman-laser orientations. The dual-atom-interferometer gyroscope is applied in\nthe measurement of the Earth rotation. The sensitivity is $1.2\\times10^{-6}$\nrad/s/$\\sqrt{Hz}$, and the long-term stability is $6.2\\times10^{-8}$ rad/s $@$\n$2000$ s.\n",
"title": "Calibration of atomic trajectories in a large-area dual-atom-interferometer gyroscope"
}
| null | null | null | null | true | null |
13409
| null |
Default
| null | null |
null |
{
"abstract": " We study the Laplacian in a smooth bounded domain, with a varying Robin\nboundary condition singular at one point. The associated quadratic form is not\nsemi-bounded from below, and the corresponding Laplacian is not self-adjoint,\nit has the residual spectrum covering the whole complex plane. We describe its\nself-adjoint extensions and exhibit a physically relevant skew-symmetric one.\nWe approximate the boundary condition, giving rise to a family of self-adjoint\noperators, and we describe their eigenvalues by the method of matched\nasymptotic expansions. These eigenvalues acquire a strange behaviour when the\nsmall perturbation parameter $\\varepsilon>0$ tends to zero, namely they become\nalmost periodic in the logarithmic scale $|\\ln \\epsilon|$ and, in this way,\n\"wander\" along the real axis at a speed $O(\\eps^{-1})$.\n",
"title": "Self-adjoint and skew-symmetric extensions of the Laplacian with singular Robin boundary condition"
}
| null | null | null | null | true | null |
13410
| null |
Default
| null | null |
null |
{
"abstract": " Geometrical and topological phases play a fundamental role in quantum theory.\nGeometric phases have been proposed as a tool for implementing unitary gates\nfor quantum computation. A fractional topological phase has been recently\ndiscovered for bipartite systems. The dimension of the Hilbert space determines\nthe topological phase of entangled qudits under local unitary operations. Here\nwe investigate fractional topological phases acquired by photonic entangled\nqudits. Photon pairs prepared as spatial qudits are operated inside a Sagnac\ninterferometer and the two-photon interference pattern reveals the topological\nphase as fringes shifts when local operations are performed. Dimensions $d = 2,\n3$ and $4$ were tested, showing the expected theoretical values.\n",
"title": "Experimental observation of fractional topological phases with photonic qudits"
}
| null | null | null | null | true | null |
13411
| null |
Default
| null | null |
null |
{
"abstract": " We study the homotopy groups of generic leaves of logarithmic foliations on\ncomplex projective manifolds. We exhibit a relation between the homotopy groups\nof a generic leaf and of the complement of the polar divisor of the logarithmic\nfoliation.\n",
"title": "Homotopy groups of generic leaves of logarithmic foliations"
}
| null | null | null | null | true | null |
13412
| null |
Default
| null | null |
null |
{
"abstract": " We exhibit relations between van Kampen-Flores, Conway-Gordon-Sachs and Radon\ntheorems, by presenting direct proofs of some implications between them. The\nkey idea is an interesting relation between the van Kampen and the\nConway-Gordon-Sachs numbers for restrictions of a map of $(d+2)$-simplex to\n$\\mathbb R^d$ to the $(d+1)$-face and to the $[d/2]$-skeleton.\n",
"title": "On van Kampen-Flores, Conway-Gordon-Sachs and Radon theorems"
}
| null | null | null | null | true | null |
13413
| null |
Default
| null | null |
null |
{
"abstract": " Let $\\Lambda = \\{\\lambda_{k}\\}$ denote a sequence of complex numbers and\nassume that that the counting function $#\\{\\lambda_{k} \\in \\Lambda : |\n\\lambda_{k}| < T\\} =O(T^{n})$ for some integer $n$. From Hadamard's theorem, we\ncan construct an entire function $f$ of order at most $n$ such that $\\Lambda$\nis the divisor $f$. In this article we prove, under reasonably general\nconditions, that the superzeta function $\\Z_{f}(s,z)$ associated to $\\Lambda$\nadmits a meromorphic continuation. Furthermore, we describe the relation\nbetween the regularized product of the sequence $z-\\Lambda$ and the function\n$f$ as constructed as a Weierstrass product. In the case $f$ admits a Dirichlet\nseries expansion in some right half-plane, we derive the meromorphic\ncontinuation in $s$ of $\\Z_{f}(s,z)$ as an integral transform of $f'/f$. We\napply these results to obtain superzeta product evaluations of Selberg zeta\nfunction associated to finite volume hyperbolic manifolds with cusps.\n",
"title": "Superzeta functions, regularized products, and the Selberg zeta function on hyperbolic manifolds with cusps"
}
| null | null | null | null | true | null |
13414
| null |
Default
| null | null |
null |
{
"abstract": " We investigate fine Selmer groups for elliptic curves and for Galois\nrepresentations over a number field. More specifically, we discuss Conjecture\nA, which states that the fine Selmer group of an elliptic curve over the\ncyclotomic extension is a finitely generated $\\mathbb{Z}_p$-module. The\nrelationship between this conjecture and Iwasawa's classical $\\mu=0$ conjecture\nis clarified. We also present some partial results towards the question whether\nConjecture A is invariant under isogenies.\n",
"title": "Fine Selmer Groups and Isogeny Invariance"
}
| null | null |
[
"Mathematics"
] | null | true | null |
13415
| null |
Validated
| null | null |
null |
{
"abstract": " We introduce a new feature map for barcodes that arise in persistent homology\ncomputation. The main idea is to first realize each barcode as a path in a\nconvenient vector space, and to then compute its path signature which takes\nvalues in the tensor algebra of that vector space. The composition of these two\noperations - barcode to path, path to tensor series - results in a feature map\nthat has several desirable properties for statistical learning, such as\nuniversality and characteristicness, and achieves state-of-the-art results on\ncommon classification benchmarks.\n",
"title": "Persistence paths and signature features in topological data analysis"
}
| null | null |
[
"Statistics"
] | null | true | null |
13416
| null |
Validated
| null | null |
null |
{
"abstract": " Our aim in this paper is to derive several new integral representations of\nthe Fox-Wright functions. In particular, we give new Laplace and Stieltjes\ntransform for this special functions under a special restriction on parameters.\nFrom the positivity conditions for the weight in these representations, we\nfound sufficient conditions to be imposed on the parameters of the Fox-Wright\nfunctions that it be completely monotonic. As applications, we derive a class\nof function related to the Fox H-functions is positive definite and an\ninvestigation of a class of the Fox H-function is non-negative. Moreover, we\nextended the Luke's inequalities and we establish a new Turán type\ninequalities for the Fox-Wright function. Finally, by appealing to each of the\nLuke's inequalities, two sets of two-sided bounding inequalities for the\ngeneralized Mathieu's type series are proved.\n",
"title": "New Integral representations for the Fox-Wright functions and its applications"
}
| null | null | null | null | true | null |
13417
| null |
Default
| null | null |
null |
{
"abstract": " During reactive transport modeling, the computational cost associated with\nchemical reaction calculations is often 10-100 times higher than that of\ntransport calculations. Most of these costs results from chemical equilibrium\ncalculations that are performed at least once in every mesh cell and at every\ntime step of the simulation. Calculating chemical equilibrium is an iterative\nprocess, where each iteration is in general so computationally expensive that\neven if every calculation converged in a single iteration, the resulting\nspeedup would not be significant. Thus, rather than proposing a fast-converging\nnumerical method for solving chemical equilibrium equations, we present a\nmachine learning method that enables new equilibrium states to be quickly and\naccurately estimated, whenever a previous equilibrium calculation with similar\ninput conditions has been performed. We demonstrate the use of this smart\nchemical equilibrium method in a reactive transport modeling example and show\nthat, even at early simulation times, the majority of all equilibrium\ncalculations are quickly predicted and, after some time steps, the\nmachine-learning-accelerated chemical solver has been fully trained to rapidly\nperform all subsequent equilibrium calculations, resulting in speedups of\nalmost two orders of magnitude. We remark that our new on-demand machine\nlearning method can be applied to any case in which a massive number of\nsequential/parallel evaluations of a computationally expensive function $f$\nneeds to be done, $y=f(x)$. We remark, that, in contrast to traditional machine\nlearning algorithms, our on-demand training approach does not require a\nstatistics-based training phase before the actual simulation of interest\ncommences. The introduced on-demand training scheme requires, however, the\nfirst-order derivatives $\\partial f/\\partial x$ for later smart predictions.\n",
"title": "Ultra-Fast Reactive Transport Simulations When Chemical Reactions Meet Machine Learning: Chemical Equilibrium"
}
| null | null | null | null | true | null |
13418
| null |
Default
| null | null |
null |
{
"abstract": " A pseudo-edge graph of a convex polyhedron K is a 3-connected embedded graph\nin K whose vertices coincide with those of K, whose edges are distance\nminimizing geodesics, and whose faces are convex. We construct a convex\npolyhedron K in Euclidean 3-space with a pseudo-edge graph E with respect to\nwhich K is not unfoldable. The proof is based on a result of Pogorelov on\nconvex caps with prescribed curvature, and an unfoldability criterion for\nalmost flat convex caps due to Tarasov. Our example, which has 340 vertices,\nsignificantly simplifies an earlier construction by Tarasov, and confirms that\nDurer's conjecture does not hold for pseudo-edge unfoldings.\n",
"title": "Pseudo-edge unfoldings of convex polyhedra"
}
| null | null | null | null | true | null |
13419
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we focus on learning structure-aware document representations\nfrom data without recourse to a discourse parser or additional annotations.\nDrawing inspiration from recent efforts to empower neural networks with a\nstructural bias, we propose a model that can encode a document while\nautomatically inducing rich structural dependencies. Specifically, we embed a\ndifferentiable non-projective parsing algorithm into a neural model and use\nattention mechanisms to incorporate the structural biases. Experimental\nevaluation across different tasks and datasets shows that the proposed model\nachieves state-of-the-art results on document modeling tasks while inducing\nintermediate structures which are both interpretable and meaningful.\n",
"title": "Learning Structured Text Representations"
}
| null | null | null | null | true | null |
13420
| null |
Default
| null | null |
null |
{
"abstract": " The Italian National Institute for Statistics regularly provides estimates of\nunemployment indicators using data from the Labor Force Survey. However, direct\nestimates of unemployment incidence cannot be released for Local Labor Market\nAreas. These are unplanned domains defined as clusters of municipalities; many\nare out-of-sample areas and the majority is characterized by a small sample\nsize, which render direct estimates inadequate. The Empirical Best Predictor\nrepresents an appropriate, model-based, alternative. However, for non-Gaussian\nresponses, its computation and the computation of the analytic approximation to\nits Mean Squared Error require the solution of (possibly) multiple integrals\nthat, generally, have not a closed form. To solve the issue, Monte Carlo\nmethods and parametric bootstrap are common choices, even though the\ncomputational burden is a non trivial task. In this paper, we propose a\nSemi-Parametric Empirical Best Predictor for a (possibly) non-linear mixed\neffect model by leaving the distribution of the area-specific random effects\nunspecified and estimating it from the observed data. This approach is known to\nlead to a discrete mixing distribution which helps avoid unverifiable\nparametric assumptions and heavy integral approximations. We also derive a\nsecond-order, bias-corrected, analytic approximation to the corresponding Mean\nSquared Error. Finite sample properties of the proposed approach are tested via\na large scale simulation study. Furthermore, the proposal is applied to\nunit-level data from the 2012 Italian Labor Force Survey to estimate\nunemployment incidence for 611 Local Labor Market Areas using auxiliary\ninformation from administrative registers and the 2011 Census.\n",
"title": "Semi-Parametric Empirical Best Prediction for small area estimation of unemployment indicators"
}
| null | null | null | null | true | null |
13421
| null |
Default
| null | null |
null |
{
"abstract": " The inertialess fluid-structure interactions of active and passive\ninextensible filaments and slender- rods are ubiquitous in nature, from the\ndynamics of semi-flexible polymers and cytoskeletal filaments to cellular\nmechanics and flagella. The coupling between the geometry of deformation and\nthe phys- ical interaction governing the dynamics of bio-filaments is complex.\nGoverning equations negotiate elastohydrodynamical interactions with\nnon-holonomic constraints arising from the filament inex- tensibility. Such\nelastohydrodynamic systems are structurally convoluted, prone to numerical\nerros, thus requiring penalization methods and high-order spatiotemporal\npropagators. The asymptotic coarse-graining formulation presented here exploits\nthe momentum balance in the asymptotic limit of small rod-like elements which\nare integrated semi-analytically. This greatly simplifies the elas-\ntohydrodynamic interactions and overcomes previous numerical instability. The\nresulting matricial system is straightforward and intuitive to implement, and\nallows for a fast and efficient computation, over than a hundred times faster\nthan previous schemes. Only basic knowledge of systems of linear equations is\nrequired, and implementation achieved with any solver of choice.\nGeneralisations for complex interaction of multiple rods, Brownian polymer\ndynamics, active filaments and non-local hydrodynamics are also\nstraightforward. We demonstrate these in four examples commonly found in\nbiological systems, including the dynamics of filaments and flagella. Three of\nthese systems are novel in the literature. We additionally provide a Matlab\ncode that can be used as a basis for further generalisations.\n",
"title": "The asymptotic coarse-graining formulation of slender-rods, bio-filaments and flagella"
}
| null | null | null | null | true | null |
13422
| null |
Default
| null | null |
null |
{
"abstract": " Encouraged by recent studies on the performance of tidal turbine arrays, we\nextend the classical momentum actuator disc theory to include the free surface\neffects and allow the vertical arrangement of turbines. Most existing\nliteratures concern one dimensional arrays with single turbine in the vertical\ndirection, while the arrays in this work are two dimensional (with turbines in\nboth the vertical and lateral directions) and also partially block the channel\nwhich width is far larger than height. The vertical mixing of array scale flow\nis assumed to take place much faster than lateral one. This assumption has been\nverified by numerical simulations. Fixing the total turbine area and utilized\nwidth, the comparison between two-dimensional and traditional one-dimensional\narrays is investigated. The results suggest that the two dimensional\narrangements of smaller turbines are preferred to one dimensional arrays from\nboth the power coefficient and efficiency perspectives. When channel dynamics\nare considered, the power increase would be partly offset according to the\nparameters of the channel but the optimal arrangement is unchangeable.\nFurthermore, we consider how to arrange finite number of turbines in a channel.\nIt is shown that an optimal distribution of turbines in two directions is\nfound. Finally, the scenario of arranging turbines in infinite flow, which is\nthe limiting condition of small blockages, is analysed. A new maximum power\ncoefficient 0.869 occurs when $Fr=0.2$, greatly increasing the peak power\ncompared with existing results.\n",
"title": "Performance of two-dimensional tidal turbine arrays in free surface flow"
}
| null | null | null | null | true | null |
13423
| null |
Default
| null | null |
null |
{
"abstract": " Elicitability is a property of $\\mathbb{R}^k$-valued functionals defined on a\nset of distribution functions. These functionals represent statistical\nproperties of a distribution, for instance its mean, variance, or median. They\nare called elicitable if there exists a scoring function such that the expected\nscore under a distribution takes its unique minimum at the functional value of\nthis distribution. If such a scoring function exists, it is called strictly\nconsistent for the functional. Motivated by the recent findings of Fissler and\nZiegel concerning higher order elicitability, this thesis reviews the most\nimportant results, examples, and applications which are found in the relevant\nliterature. Moreover, we also contribute our own examples and findings in order\nto give the reader a well-founded overview of the topic as well as of the most\nused tools and techniques. We include necessary and sufficient conditions for\nstrictly consistent scoring functions, several elicitable as well as\nnon-elicitable functionals and the use of elicitability in forecast comparison,\nregression, and estimation. Special emphasis is placed on quantitative risk\nmanagement and the result that Value at Risk and Expected Shortfall are jointly\nelicitable.\n",
"title": "Elicitability and its Application in Risk Management"
}
| null | null | null | null | true | null |
13424
| null |
Default
| null | null |
null |
{
"abstract": " A system of interacting Brownian particles subject to short-range repulsive\npotentials is considered. A continuum description in the form of a nonlinear\ndiffusion equation is derived systematically in the dilute limit using the\nmethod of matched asymptotic expansions. Numerical simulations are performed to\ncompare the results of the model with those of the commonly used mean-field and\nKirkwood-superposition approximations, as well as with Monte Carlo simulation\nof the stochastic particle system, for various interaction potentials. Our\napproach works best for very repulsive short-range potentials, while the\nmean-field approximation is suitable for long-range interactions. The Kirkwood\nsuperposition approximation provides an accurate description for both short-\nand long-range potentials, but is considerably more computationally intensive.\n",
"title": "Diffusion of particles with short-range interactions"
}
| null | null | null | null | true | null |
13425
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we first present a Birman-Murakami-Wenzl type algebra for every\nCoxeter system of rank 2 (corresponding to dihedral groups). We prove they have\nsemisimple for generic parameters, and having natural cellular structures. And\nclasscify their irreducible representations. Among them there is one serving as\na generalization of the Lawrence-Krammer representation with quite neat shape\nand the \"correct\" dimension. We conjecture they are isomorphic to the\ngeneralized Lawrence-Krammer representaions defined by I.Marin as monodromy of\ncertain KZ connections. We prove these representations are irreducible for\ngeneric parameters, and find a quite neat invariant bilinear form on them.\nBased on above constructions for rank 2, we introduce a Birman-Murakami-Wenzl\ntype algebra for an arbitrary Coxeter system. For every Coxeter system, the\nintroduced algebra is a quotient of group algebra of the Artin group\n(associated with this Coxeter system), having the corresponding Hecke algebra\nas a quotient. The simple generators of the Artin group have degree 3\nannihiating polynomials in this algebra.\n",
"title": "Birman-Murakami-Wenzl type algebras for arbitrary Coxeter systems"
}
| null | null |
[
"Mathematics"
] | null | true | null |
13426
| null |
Validated
| null | null |
null |
{
"abstract": " In recent era prediction of enzyme class from an unknown protein is one of\nthe challenging tasks in bioinformatics. Day to day the number of proteins is\nincreases as result the prediction of enzyme class gives a new opportunity to\nbioinformatics scholars. The prime objective of this article is to implement\nthe machine learning classification technique for feature selection and\npredictions also find out an appropriate classification technique for function\nprediction. In this article the seven different classification technique like\nCRT, QUEST, CHAID, C5.0, ANN (Artificial Neural Network), SVM and Bayesian has\nbeen implemented on 4368 protein data that has been extracted from UniprotKB\ndatabank and categories into six different class. The proteins data is high\ndimensional sequence data and contain a maximum of 48 features.To manipulate\nthe high dimensional sequential protein data with different classification\ntechnique, the SPSS has been used as an experimental tool. Different\nclassification techniques give different results for every model and shows that\nthe data are imbalanced for class C4, C5 and C6. The imbalanced data affect the\nperformance of model. In these three classes the precision and recall value is\nvery less or negligible. The experimental results highlight that the C5.0\nclassification technique accuracy is more suited for protein feature\nclassification and predictions. The C5.0 classification technique gives 95.56%\naccuracy and also gives high precision and recall value. Finally, we conclude\nthat the features that is selected can be used for function prediction.\n",
"title": "Protein Classification using Machine Learning and Statistical Techniques: A Comparative Analysis"
}
| null | null | null | null | true | null |
13427
| null |
Default
| null | null |
null |
{
"abstract": " The origin of population-scale coordination has puzzled philosophers and\nscientists for centuries. Recently, game theory, evolutionary approaches and\ncomplex systems science have provided quantitative insights on the mechanisms\nof social consensus. However, the literature is vast and widely scattered\nacross fields, making it hard for the single researcher to navigate it. This\nshort review aims to provide a compact overview of the main dimensions over\nwhich the debate has unfolded and to discuss some representative examples. It\nfocuses on those situations in which consensus emerges 'spontaneously' in\nabsence of centralised institutions and covers topic that include the\nmacroscopic consequences of the different microscopic rules of behavioural\ncontagion, the role of social networks, and the mechanisms that prevent the\nformation of a consensus or alter it after it has emerged. Special attention is\ndevoted to the recent wave of experiments on the emergence of consensus in\nsocial systems.\n",
"title": "The Emergence of Consensus: A Primer"
}
| null | null | null | null | true | null |
13428
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we present ProSLAM, a lightweight stereo visual SLAM system\ndesigned with simplicity in mind. Our work stems from the experience gathered\nby the authors while teaching SLAM to students and aims at providing a highly\nmodular system that can be easily implemented and understood. Rather than\nfocusing on the well known mathematical aspects of Stereo Visual SLAM, in this\nwork we highlight the data structures and the algorithmic aspects that one\nneeds to tackle during the design of such a system. We implemented ProSLAM\nusing the C++ programming language in combination with a minimal set of well\nknown used external libraries. In addition to an open source implementation, we\nprovide several code snippets that address the core aspects of our approach\ndirectly in this paper. The results of a thorough validation performed on\nstandard benchmark datasets show that our approach achieves accuracy comparable\nto state of the art methods, while requiring substantially less computational\nresources.\n",
"title": "ProSLAM: Graph SLAM from a Programmer's Perspective"
}
| null | null | null | null | true | null |
13429
| null |
Default
| null | null |
null |
{
"abstract": " We introduce right generating sets, Cayley graphs, growth functions, types\nand rates, and isoperimetric constants for left homogeneous spaces equipped\nwith coordinate systems; characterise right amenable finitely right generated\nleft homogeneous spaces with finite stabilisers as those whose isoperimetric\nconstant is $0$; and prove that finitely right generated left homogeneous\nspaces with finite stabilisers of sub-exponential growth are right amenable, in\nparticular, quotient sets of groups of sub-exponential growth by finite\nsubgroups are right amenable.\n",
"title": "Right Amenability And Growth Of Finitely Right Generated Left Group Sets"
}
| null | null |
[
"Mathematics"
] | null | true | null |
13430
| null |
Validated
| null | null |
null |
{
"abstract": " A branch flow model (BFM) is used to formulate the AC power flow in general\nnetworks. For each branch/line, the BFM contains a nonconvex quadratic\nequality. A mathematical formulation of its convex hull is proposed, which is\nthe tightest convex relaxation of this quadratic equation. The convex hull\nformulation consists of a second order cone inequality and a linear inequality\nwithin the physical bounds of power flows. The convex hull formulation is\nanalytically proved and geometrically validated. An optimal scheduling problem\nof distributed energy storage (DES) in radial distribution systems with high\npenetration of photovoltaic resources is investigated in this paper. To capture\nthe performance of both the battery and converter, a second-order DES model is\nproposed. Following the convex hull of the quadratic branch flow equation, the\nconvex hull formulation of the nonconvex constraint in the DES model is also\nderived. The proposed convex hull models are used to generate a tight convex\nrelaxation of the DES optimal scheduling (DESOS) problem. The proposed approach\nis tested on several radial systems. A discussion on the extension to meshed\nnetworks is provided.\n",
"title": "Convex Hull of the Quadratic Branch AC Power Flow Equations and Its Application in Radial Distribution Networks"
}
| null | null |
[
"Mathematics"
] | null | true | null |
13431
| null |
Validated
| null | null |
null |
{
"abstract": " Transfer learning aims to faciliate learning tasks in a label-scarce target\ndomain by leveraging knowledge from a related source domain with plenty of\nlabeled data. Often times we may have multiple domains with little or no\nlabeled data as targets waiting to be solved. Most existing efforts tackle\ntarget domains separately by modeling the `source-target' pairs without\nexploring the relatedness between them, which would cause loss of crucial\ninformation, thus failing to achieve optimal capability of knowledge transfer.\nIn this paper, we propose a novel and effective approach called Multi-Relevance\nTransfer Learning (MRTL) for this purpose, which can simultaneously transfer\ndifferent knowledge from the source and exploits the shared common latent\nfactors between target domains. Specifically, we formulate the problem as an\noptimization task based on a collective nonnegative matrix tri-factorization\nframework. The proposed approach achieves both source-target transfer and\ntarget-target leveraging by sharing multiple decomposed latent subspaces.\nFurther, an alternative minimization learning algorithm is developed with\nconvergence guarantee. Empirical study validates the performance and\neffectiveness of MRTL compared to the state-of-the-art methods.\n",
"title": "Multi-Relevance Transfer Learning"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
13432
| null |
Validated
| null | null |
null |
{
"abstract": " Group discussions are a way for individuals to exchange ideas and arguments\nin order to reach better decisions than they could on their own. One of the\npremises of productive discussions is that better solutions will prevail, and\nthat the idea selection process is mediated by the (relative) competence of the\nindividuals involved. However, since people may not know their actual\ncompetence on a new task, their behavior is influenced by their self-estimated\ncompetence --- that is, their confidence --- which can be misaligned with their\nactual competence.\nOur goal in this work is to understand the effects of confidence-competence\nmisalignment on the dynamics and outcomes of discussions. To this end, we\ndesign a large-scale natural setting, in the form of an online team-based\ngeography game, that allows us to disentangle confidence from competence and\nthus separate their effects.\nWe find that in task-oriented discussions, the more-confident individuals\nhave a larger impact on the group's decisions even when these individuals are\nat the same level of competence as their teammates. Furthermore, this\nunjustified role of confidence in the decision-making process often leads teams\nto under-perform. We explore this phenomenon by investigating the effects of\nconfidence on conversational dynamics.\n",
"title": "When confidence and competence collide: Effects on online decision-making discussions"
}
| null | null | null | null | true | null |
13433
| null |
Default
| null | null |
null |
{
"abstract": " In this short note, using results of Bourgain, Fremlin, and Talagrand\n\\cite{BFT}, we show that for a countable structure $M$, a saturated elementary\nextension $M^*$ of $M$ and a formula $\\phi(x,y)$ the following are equivalent:\n(i) $\\phi(x,y)$ is NIP on $M$ (in the sense of Definition 2.1).\n(ii) Whenever $p(x)\\in S_\\phi(M^*)$ is finitely satisfiable in $M$ then it is\nBaire 1 definable over $M$ (in sense of Definition 2.5).\n",
"title": "NIP formulas and Baire 1 definability"
}
| null | null | null | null | true | null |
13434
| null |
Default
| null | null |
null |
{
"abstract": " In this article, we have modeled mortality rates of Peruvian female and male\npopulations during the period of 1950-2017 using the Lee-Carter (LC) model. The\nstochastic mortality model was introduced by Lee and Carter (1992) and has been\nused by many authors for fitting and forecasting the human mortality rates. The\nSingular Value Decomposition (SVD) approach is used for estimation of the\nparameters of the LC model. Utilizing the best fitted auto regressive\nintegrated moving average (ARIMA) model we forecast the values of the time\ndependent parameter of the LC model for the next thirty years. The forecasted\nvalues of life expectancy at different age group with $95\\%$ confidence\nintervals are also reported for the next thirty years. In this research we use\nthe data, obtained from the Peruvian National Institute of Statistics (INEI).\n",
"title": "Lee-Carter method for forecasting mortality for Peruvian Population"
}
| null | null | null | null | true | null |
13435
| null |
Default
| null | null |
null |
{
"abstract": " We propose a novel approach to parameter estimation for simulator-based\nstatistical models with intractable likelihood. Our proposed method involves\nrecursive application of kernel ABC and kernel herding to the same observed\ndata. We provide a theoretical explanation regarding why the approach works,\nshowing (for the population setting) that, under a certain assumption, point\nestimates obtained with this method converge to the true parameter, as\nrecursion proceeds. We have conducted a variety of numerical experiments,\nincluding parameter estimation for a real-world pedestrian flow simulator, and\nshow that in most cases our method outperforms existing approaches.\n",
"title": "Kernel Recursive ABC: Point Estimation with Intractable Likelihood"
}
| null | null | null | null | true | null |
13436
| null |
Default
| null | null |
null |
{
"abstract": " Several Prolog implementations include a facility for tabling, an alternative\nresolution strategy which uses memoisation to avoid redundant duplication of\ncomputations. Until relatively recently, tabling has required either low-level\nsupport in the underlying Prolog engine, or extensive program transormation (de\nGuzman et al., 2008). An alternative approach is to augment Prolog with low\nlevel support for continuation capturing control operators, particularly\ndelimited continuations, which have been investigated in the field of\nfunctional programming and found to be capable of supporting a wide variety of\ncomputational effects within an otherwise declarative language.\nThis technical report describes an implementation of tabling in SWI Prolog\nbased on delimited control operators for Prolog recently introduced by\nSchrijvers et al. (2013). In comparison with a previous implementation of\ntabling for SWI Prolog using delimited control (Desouter et al., 2015), this\napproach, based on the functional memoising parser combinators of Johnson\n(1995), stays closer to the declarative core of Prolog, requires less code, and\nis able to deliver solutions from systems of tabled predicates incrementally\n(as opposed to finding all solutions before delivering any to the rest of the\nprogram).\nA collection of benchmarks shows that a small number of carefully targeted\noptimisations yields performance within a factor of about 2 of the optimised\nversion of Desouter et al.'s system currently included in SWI Prolog.\n",
"title": "More declarative tabling in Prolog using multi-prompt delimited control"
}
| null | null | null | null | true | null |
13437
| null |
Default
| null | null |
null |
{
"abstract": " We introduce the notion of a dynamical topological order parameter (DTOP)\nthat characterises dynamical quantum phase transitions (DQPTs) occurring in the\nsubsequent temporal evolution of \"two dimensional\" closed quantum systems,\nfollowing a quench (or ramping) of a parameter of the Hamiltonian, {which\ngeneralizes the notion of DTOP introduced in Budich and Heyl, Phys. Rev. B 93,\n085416 (2016) for one-dimensional situations}. This DTOP is obtained from the\n\"gauge-invariant\" Pancharatnam phase extracted from the Loschmidt overlap,\ni.e., the modulus of the overlap between the initially prepared state and its\ntime evolved counterpart reached following a temporal evolution generated by\nthe time-independent final Hamiltonian. This generic proposal is illustrated\nconsidering DQPTs occurring in the subsequent temporal evolution following a\nsudden quench of the staggered mass of the topological Haldane model on a\nhexagonal lattice where it stays fixed to zero or unity and makes a\ndiscontinuous jump between these two values at critical times at which DQPTs\noccur.\n",
"title": "Emergent topology and dynamical quantum phase transitions in two-dimensional closed quantum systems"
}
| null | null | null | null | true | null |
13438
| null |
Default
| null | null |
null |
{
"abstract": " In this article we present pictorially the foundation of differential\ngeometry which is a crucial tool for multiple areas of physics, notably general\nand special relativity, but also mechanics, thermodynamics and solving\ndifferential equations. As all the concepts are presented as pictures, there\nare no equations in this article. As such this article may be read by\npre-university students who enjoy physics, mathematics and geometry. However it\nwill also greatly aid the intuition of an undergraduate and masters students,\nlearning general relativity and similar courses. It concentrates on the tools\nneeded to understand Maxwell's equations thus leading to the goal of presenting\nMaxwell's equations as 3 pictures.\n",
"title": "A pictorial introduction to differential geometry, leading to Maxwell's equations as three pictures"
}
| null | null | null | null | true | null |
13439
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we study a non strictly systems of conservation law by\nstochastic perturbation. We show the existence and uniqueness of the solution.\nWe do not assume that $BV$-regularity for the initial conditions. The proofs\nare based on the concept of entropy solution and in the characteristics method\n(in the influence of noise). This is the first result on the regularization by\nnoise in hyperbolic systems of conservation law.\n",
"title": "Regularization by noise in (2x 2) hyperbolic systems of conservation law"
}
| null | null | null | null | true | null |
13440
| null |
Default
| null | null |
null |
{
"abstract": " The one-dimensional symmetric exclusion process, the simplest interacting\nparticle process, is a lattice-gas made of particles that hop symmetrically on\na discrete line respecting hard-core exclusion. The system is prepared on the\ninfinite lattice with a step initial profile with average densities $\\rho_{+}$\nand $\\rho_{-}$ on the right and on the left of the origin. When $\\rho_{+} =\n\\rho_{-}$, the gas is at equilibrium and undergoes stationary fluctuations.\nWhen these densities are unequal, the gas is out of equilibrium and will remain\nso forever. A tracer, or a tagged particle, is initially located at the\nboundary between the two domains; its position $X_t$ is a random observable in\ntime, that carries information on the non-equilibrium dynamics of the whole\nsystem. We derive an exact formula for the cumulant generating function and the\nlarge deviation function of $X_t$, in the long time limit, and deduce the full\nstatistical properties of the tracer's position. The equilibrium fluctuations\nof the tracer's position, when the density is uniform, are obtained as an\nimportant special case.\n",
"title": "Large deviations of a tracer in the symmetric exclusion process"
}
| null | null | null | null | true | null |
13441
| null |
Default
| null | null |
null |
{
"abstract": " This paper classifies the equivalence classes of irreducible unitary\nrepresentations with nonvanishing Dirac cohomology for complex $E_6$. This is\nachieved by using our finiteness result, and by improving the computing method.\n",
"title": "Unitary Representations with non-zero Dirac cohomology for complex $E_6$"
}
| null | null | null | null | true | null |
13442
| null |
Default
| null | null |
null |
{
"abstract": " After a short review of the classical Lie theorem, a finite dimensional Lie\nalgebra of vector fields is considered and the most general conditions under\nwhich the integral curves of one of the fields can be obtained by quadratures\nin a prescribed way will be discussed, determining also the number of\nquadratures needed to integrate the system. The theory will be illustrated with\nexamples andbn an extension of the theorem where the Lie algebras are replaced\nby some distributions will also be presented.\n",
"title": "Revisiting Lie integrability by quadratures from a geometric perspective"
}
| null | null | null | null | true | null |
13443
| null |
Default
| null | null |
null |
{
"abstract": " Intersubband (ISB) polarons result from the interaction of an ISB transition\nand the longitudinal optical (LO) phonons in a semiconductor quantum well (QW).\nTheir observation requires a very dense two dimensional electron gas (2DEG) in\nthe QW and a polar or highly ionic semiconductor. Here we show that in\nZnO/MgZnO QWs the strength of such a coupling can be as high as 1.5 times the\nLO-phonon frequency due to the very dense 2DEG achieved and the large\ndifference between the static and high-frequency dielectric constants in ZnO.\nThe ISB polaron is observed optically in multiple QW structures with 2DEG\ndensities ranging from $5\\times 10^{12}$ to $5\\times 10^{13}$ cm$^{-2}$, where\nan unprecedented regime is reached in which the frequency of the upper ISB\npolaron branch is three times larger than that of the bare ISB transition. This\nstudy opens new prospects to the exploitation of oxides in phenomena happening\nin the ultrastrong coupling regime.\n",
"title": "Intersubband polarons in oxides"
}
| null | null | null | null | true | null |
13444
| null |
Default
| null | null |
null |
{
"abstract": " It is well known that Markov chain Monte Carlo (MCMC) methods scale poorly\nwith dataset size. A popular class of methods for solving this issue is\nstochastic gradient MCMC. These methods use a noisy estimate of the gradient of\nthe log posterior, which reduces the per iteration computational cost of the\nalgorithm. Despite this, there are a number of results suggesting that\nstochastic gradient Langevin dynamics (SGLD), probably the most popular of\nthese methods, still has computational cost proportional to the dataset size.\nWe suggest an alternative log posterior gradient estimate for stochastic\ngradient MCMC, which uses control variates to reduce the variance. We analyse\nSGLD using this gradient estimate, and show that, under log-concavity\nassumptions on the target distribution, the computational cost required for a\ngiven level of accuracy is independent of the dataset size. Next we show that a\ndifferent control variate technique, known as zero variance control variates\ncan be applied to SGMCMC algorithms for free. This post-processing step\nimproves the inference of the algorithm by reducing the variance of the MCMC\noutput. Zero variance control variates rely on the gradient of the log\nposterior; we explore how the variance reduction is affected by replacing this\nwith the noisy gradient estimate calculated by SGMCMC.\n",
"title": "Control Variates for Stochastic Gradient MCMC"
}
| null | null | null | null | true | null |
13445
| null |
Default
| null | null |
null |
{
"abstract": " One of the major challenges in Minimally Invasive Surgery (MIS) such as\nlaparoscopy is the lack of depth perception. In recent years, laparoscopic\nscene tracking and surface reconstruction has been a focus of investigation to\nprovide rich additional information to aid the surgical process and compensate\nfor the depth perception issue. However, robust 3D surface reconstruction and\naugmented reality with depth perception on the reconstructed scene are yet to\nbe reported. This paper presents our work in this area. First, we adopt a\nstate-of-the-art visual simultaneous localization and mapping (SLAM) framework\n- ORB-SLAM - and extend the algorithm for use in MIS scenes for reliable\nendoscopic camera tracking and salient point mapping. We then develop a robust\nglobal 3D surface reconstruction frame- work based on the sparse point clouds\nextracted from the SLAM framework. Our approach is to combine an outlier\nremoval filter within a Moving Least Squares smoothing algorithm and then\nemploy Poisson surface reconstruction to obtain smooth surfaces from the\nunstructured sparse point cloud. Our proposed method has been quantitatively\nevaluated compared with ground-truth camera trajectories and the organ model\nsurface we used to render the synthetic simulation videos. In vivo laparoscopic\nvideos used in the tests have demonstrated the robustness and accuracy of our\nproposed framework on both camera tracking and surface reconstruction,\nillustrating the potential of our algorithm for depth augmentation and\ndepth-corrected augmented reality in MIS with monocular endoscopes.\n",
"title": "Augmented Reality for Depth Cues in Monocular Minimally Invasive Surgery"
}
| null | null | null | null | true | null |
13446
| null |
Default
| null | null |
null |
{
"abstract": " This paper describes Task 2 of the DCASE 2018 Challenge, titled\n\"General-purpose audio tagging of Freesound content with AudioSet labels\". This\ntask was hosted on the Kaggle platform as \"Freesound General-Purpose Audio\nTagging Challenge\". The goal of the task is to build an audio tagging system\nthat can recognize the category of an audio clip from a subset of 41 diverse\ncategories drawn from the AudioSet Ontology. We present the task, the dataset\nprepared for the competition, and a baseline system.\n",
"title": "General-purpose Tagging of Freesound Audio with AudioSet Labels: Task Description, Dataset, and Baseline"
}
| null | null | null | null | true | null |
13447
| null |
Default
| null | null |
null |
{
"abstract": " The program dependence graph (PDG) represents data and control dependence\nbetween statements in a program. This paper presents an operational semantics\nof program dependence graphs. Since PDGs exclude artificial order of statements\nthat resides in sequential programs, executions of PDGs are not unique.\nHowever, we identified a class of PDGs that have unique final states of\nexecutions, called deterministic PDGs. We prove that the operational semantics\nof control flow graphs is equivalent to that of deterministic PDGs. The class\nof deterministic PDGs properly include PDGs obtained from well-structured\nprograms. Thus, our operational semantics of PDGs is more general than that of\nPDGs for well-structured programs, which are already established in literature.\n",
"title": "Semantical Equivalence of the Control Flow Graph and the Program Dependence Graph"
}
| null | null | null | null | true | null |
13448
| null |
Default
| null | null |
null |
{
"abstract": " We prove that the smallest non-trivial quotient of the mapping class group of\na connected orientable surface of genus at least 3 without punctures is\n$\\mathrm{Sp}_{2g}(2)$, thus confirming a conjecture of Zimmermann. In the\nprocess, we generalise Korkmaz's results on $\\mathbb{C}$-linear representations\nof mapping class groups to projective representations over any field.\n",
"title": "On the smallest non-trivial quotients of mapping class groups"
}
| null | null | null | null | true | null |
13449
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we present a new type of fractional operator, which is a\ngeneralization of the Caputo and Caputo--Hadamard fractional derivative\noperators. We study some properties of the operator, namely we prove that it is\nthe inverse operation of a generalized fractional integral. A relation between\nthis operator and a Riemann--Liouville type is established. We end with a\nfractional Gronwall inequality type, which is useful to compare solutions of\nfractional differential equations.\n",
"title": "A Gronwall inequality for a general Caputo fractional operator"
}
| null | null | null | null | true | null |
13450
| null |
Default
| null | null |
null |
{
"abstract": " In this research, we employ accurate time-dependent density functional\ncalculations for ultrashort laser spectroscopy of nitrogen molecule. Laser\npulses with different frequencies, intensities, and durations are applied to\nthe molecule and the resulting photoelectron spectra are analyzed. It is argued\nthat relative orientation of the molecule in the laser pulse significantly\ninfluence the orbital character of the emitted photoelectrons. Moreover, the\nduration of the laser pulse is also found to be very effective in controlling\nthe orbital resolution and intensity of photoelectrons. Angular resolved\ndistribution of photoelectrons are computed at different pulse frequencies and\nrecording times. By exponential growth of the laser pulse intensity, the\ntheoretical threshold of two photons absorption in nitrogen molecule is\ndetermined.\n",
"title": "First-principles insights into ultrashort laser spectroscopy of molecular nitrogen"
}
| null | null | null | null | true | null |
13451
| null |
Default
| null | null |
null |
{
"abstract": " In this note we prove that the Borel class of representations of 3-manifold\ngroups to PGL(n,C) is preserved under Cartan involution up to sign. For\nrepresentations to PGL(3,C) this is implied by a more general result of E.\nFalbel and Q. Wang, however our proof appears to be much shorter for that\nspecial case.\n",
"title": "Borel class and Cartan involution"
}
| null | null | null | null | true | null |
13452
| null |
Default
| null | null |
null |
{
"abstract": " In this article we make a case for a systematic application of complex\nnetwork science to study art market history and more general collection\ndynamics. We reveal social, temporal, spatial, and conceptual network\ndimensions, i.e. network node and link types, previously implicit in the Getty\nProvenance Index (GPI). As a pioneering art history database active since the\n1980s, the GPI provides online access to source material relevant for research\nin the history of collecting and art markets. Based on a subset of the GPI, we\ncharacterize an aggregate of more than 267,000 sales transactions connected to\nroughly 22,000 actors in four countries over 20 years at daily resolution from\n1801 to 1820. Striving towards a deeper understanding on multiple levels we\ndisambiguate social dynamics of buying, brokering, and selling, while observing\na general broadening of the market, where large collections are split into\nsmaller lots. Temporally, we find annual market cycles that are shifted by\ncountry and obviously favor international exchange. Spatially, we differentiate\nnear-monopolies from regions driven by competing sub-centers, while uncovering\nasymmetries of international market flux. Conceptually, we track dynamics of\nartist attribution that clearly behave like product categories in a very slow\nsupermarket. Taken together, we introduce a number of meaningful network\nperspectives dealing with historical art auction data, beyond the analysis of\nsocial networks within a single market region. The results presented here have\ninspired a Linked Open Data conversion of the GPI, which is currently in\nprocess and will allow further analysis by a broad set of researchers.\n",
"title": "Network Dimensions in the Getty Provenance Index"
}
| null | null | null | null | true | null |
13453
| null |
Default
| null | null |
null |
{
"abstract": " We estimate the maximum-order complexity of a binary sequence in terms of its\ncorrelation measures. Roughly speaking, we show that any sequence with small\ncorrelation measure up to a sufficiently large order $k$ cannot have very small\nmaximum-order complexity.\n",
"title": "Maximum-order Complexity and Correlation Measures"
}
| null | null | null | null | true | null |
13454
| null |
Default
| null | null |
null |
{
"abstract": " Recent advances in neural networks have inspired people to design hybrid\nrecommendation algorithms that can incorporate both (1) user-item interaction\ninformation and (2) content information including image, audio, and text.\nDespite their promising results, neural network-based recommendation algorithms\npose extensive computational costs, making it challenging to scale and improve\nupon. In this paper, we propose a general neural network-based recommendation\nframework, which subsumes several existing state-of-the-art recommendation\nalgorithms, and address the efficiency issue by investigating sampling\nstrategies in the stochastic gradient descent training for the framework. We\ntackle this issue by first establishing a connection between the loss functions\nand the user-item interaction bipartite graph, where the loss function terms\nare defined on links while major computation burdens are located at nodes. We\ncall this type of loss functions \"graph-based\" loss functions, for which varied\nmini-batch sampling strategies can have different computational costs. Based on\nthe insight, three novel sampling strategies are proposed, which can\nsignificantly improve the training efficiency of the proposed framework (up to\n$\\times 30$ times speedup in our experiments), as well as improving the\nrecommendation performance. Theoretical analysis is also provided for both the\ncomputational cost and the convergence. We believe the study of sampling\nstrategies have further implications on general graph-based loss functions, and\nwould also enable more research under the neural network-based recommendation\nframework.\n",
"title": "On Sampling Strategies for Neural Network-based Collaborative Filtering"
}
| null | null | null | null | true | null |
13455
| null |
Default
| null | null |
null |
{
"abstract": " We study interactions between bright matter-wave solitons which acquire\nchiral transport dynamics due to an optically-induced density-dependent gauge\npotential. Through numerical simulations, we find that the collision dynamics\nfeature several non-integrable phenomena, from inelastic collisions including\npopulation transfer and radiation losses to short-lived bound states and\nsoliton fission. An effective quasi-particle model for the interaction between\nthe solitons is derived by means of a variational approximation, which\ndemonstrates that the inelastic nature of the collision arises from a coupling\nof the gauge field to velocities of the solitons. In addition, we derive a set\nof interaction potentials which show that the influence of the gauge field\nappears as a short-range potential, that can give rise to both attractive and\nrepulsive interactions.\n",
"title": "Non-integrable dynamics of matter-wave solitons in a density-dependent gauge theory"
}
| null | null | null | null | true | null |
13456
| null |
Default
| null | null |
null |
{
"abstract": " We give geometric characterisations of patch and Lawson topologies in the\ncontext of predicative point-free topology using the constructive notion of\nlocated subset. We present the patch topology of a stably locally compact\nformal topology by a geometric theory whose models are the points of the given\ntopology that are located, and the Lawson topology of a continuous lattice by a\ngeometric theory whose models are the located subsets of the given lattice. We\nalso give a predicative presentation of the frame of perfect nuclei on a stably\nlocally compact formal topology, and show that it is essentially the same as\nour geometric presentation of the patch topology. Moreover, the construction of\nLawson topologies naturally induces a monad on the category of compact regular\nformal topologies, which is shown to be isomorphic to the Vietoris monad.\n",
"title": "Geometric theories of patch and Lawson topologies"
}
| null | null | null | null | true | null |
13457
| null |
Default
| null | null |
null |
{
"abstract": " In about four years, the National Aeronautics and Space Administration (NASA)\nwill launch a small explorer mission named the Imaging X-ray Polarimetry\nExplorer (IXPE). IXPE is a satellite dedicated to the observation of X-ray\npolarization from bright astronomical sources in the 2-8 keV energy range.\nUsing Gas Pixel Detectors (GPD), the mission will allow for the first time to\nacquire X-ray polarimetric imaging and spectroscopy of about a hundred of\nsources during its first two years of operation. Among them are the most\npowerful sources of light in the Universe: active galactic nuclei (AGN). In\nthis proceedings, we summarize the scientific exploration we aim to achieve in\nthe field of AGN using IXPE, describing the main discoveries that this new\ngeneration of X-ray polarimeters will be able to make. Among these discoveries,\nwe expect to detect indisputable signatures of strong gravity, quantifying the\namount and importance of scattering on distant cold material onto the iron\nK_alpha line observed at 6.4 keV. IXPE will also be able to probe the\nmorphology of parsec-scale AGN regions, the magnetic field strength and\ndirection in quasar jets, and, among the most important results, deliver an\nindependent measurement of the spin of black holes.\n",
"title": "Active galactic nuclei in the era of the Imaging X-ray Polarimetry Explorer"
}
| null | null |
[
"Physics"
] | null | true | null |
13458
| null |
Validated
| null | null |
null |
{
"abstract": " In the present paper we use the theory of exact completions to study\ncategorical properties of small setoids in Martin-Löf type theory and, more\ngenerally, of models of the Constructive Elementary Theory of the Category of\nSets, in terms of properties of their subcategories of choice objects (i.e.\nobjects satisfying the axiom of choice). Because of these intended\napplications, we deal with categories that lack equalisers and just have weak\nones, but whose objects can be regarded as collections of global elements. In\nthis context, we study the internal logic of the categories involved, and\nemploy this analysis to give a sufficient condition for the local cartesian\nclosure of an exact completion. Finally, we apply these results to show when an\nexact completion produces a model of CETCS.\n",
"title": "Exact completion and constructive theories of sets"
}
| null | null | null | null | true | null |
13459
| null |
Default
| null | null |
null |
{
"abstract": " Many Program Verification and Synthesis problems of interest can be modeled\ndirectly using Horn clauses and many recent advances in the CLP and CAV\ncommunities have centered around efficiently solving problems presented as Horn\nclauses.\nThe HCVS series of workshops aims to bring together researchers working in\nthe two communities of Constraint/Logic Programming (e.g., ICLP and CP),\nProgram Verification (e.g., CAV, TACAS, and VMCAI), and Automated Deduction\n(e.g., CADE, IJCAR), on the topic of Horn clause based analysis, verification,\nand synthesis.\nHorn clauses for verification and synthesis have been advocated by these\ncommunities in different times and from different perspectives and HCVS is\norganized to stimulate interaction and a fruitful exchange and integration of\nexperiences.\n",
"title": "Proceedings 5th Workshop on Horn Clauses for Verification and Synthesis"
}
| null | null | null | null | true | null |
13460
| null |
Default
| null | null |
null |
{
"abstract": " Deep neural networks excel at function approximation, yet they are typically\ntrained from scratch for each new function. On the other hand, Bayesian\nmethods, such as Gaussian Processes (GPs), exploit prior knowledge to quickly\ninfer the shape of a new function at test time. Yet GPs are computationally\nexpensive, and it can be hard to design appropriate priors. In this paper we\npropose a family of neural models, Conditional Neural Processes (CNPs), that\ncombine the benefits of both. CNPs are inspired by the flexibility of\nstochastic processes such as GPs, but are structured as neural networks and\ntrained via gradient descent. CNPs make accurate predictions after observing\nonly a handful of training data points, yet scale to complex functions and\nlarge datasets. We demonstrate the performance and versatility of the approach\non a range of canonical machine learning tasks, including regression,\nclassification and image completion.\n",
"title": "Conditional Neural Processes"
}
| null | null | null | null | true | null |
13461
| null |
Default
| null | null |
null |
{
"abstract": " Many earth science applications require data at both high spatial and\ntemporal resolution for effective monitoring of various ecosystem resources.\nDue to practical limitations in sensor design, there is often a trade-off in\ndifferent resolutions of spatio-temporal datasets and hence a single sensor\nalone cannot provide the required information. Various data fusion methods have\nbeen proposed in the literature that mainly rely on individual timesteps when\nboth datasets are available to learn a mapping between features values at\ndifferent resolutions using local relationships between pixels. Earth\nobservation data is often plagued with spatially and temporally correlated\nnoise, outliers and missing data due to atmospheric disturbances which pose a\nchallenge in learning the mapping from a local neighborhood at individual\ntimesteps. In this paper, we aim to exploit time-independent global\nrelationships between pixels for robust transfer of information across\ndifferent scales. Specifically, we propose a new framework, ORBIT (Ordering\nBased Information Transfer) that uses relative ordering constraint among pixels\nto transfer information across both time and scales. The effectiveness of the\nframework is demonstrated for global surface water monitoring using both\nsynthetic and real-world datasets.\n",
"title": "ORBIT: Ordering Based Information Transfer Across Space and Time for Global Surface Water Monitoring"
}
| null | null | null | null | true | null |
13462
| null |
Default
| null | null |
null |
{
"abstract": " Psychiatric neuroscience is increasingly aware of the need to define\npsychopathology in terms of abnormal neural computation. The central tool in\nthis endeavour is the fitting of computational models to behavioural data. The\nmost prominent example of this procedure is fitting reinforcement learning (RL)\nmodels to decision-making data collected from mentally ill and healthy subject\npopulations. These models are generative models of the decision-making data\nthemselves, and the parameters we seek to infer can be psychologically and\nneurobiologically meaningful. Currently, the gold standard approach to this\ninference procedure involves Monte-Carlo sampling, which is robust but\ncomputationally intensive---rendering additional procedures, such as\ncross-validation, impractical. Searching for point estimates of model\nparameters using optimization procedures remains a popular and interesting\noption. On a novel testbed simulating parameter estimation from a common RL\ntask, we investigated the effects of smooth vs. boundary constraints on\nparameter estimation using interior point and deterministic direct search\nalgorithms for optimization. Ultimately, we show that the use of boundary\nconstraints can lead to substantial truncation effects. Our results discourage\nthe use of boundary constraints for these applications.\n",
"title": "The Importance of Constraint Smoothness for Parameter Estimation in Computational Cognitive Modeling"
}
| null | null | null | null | true | null |
13463
| null |
Default
| null | null |
null |
{
"abstract": " The maximum speed at which a liquid can wet a solid is limited by the need to\ndisplace gas lubrication films in front of the moving contact line. The\ncharacteristic height of these films is often comparable to the mean free path\nin the gas so that hydrodynamic models do not adequately describe the flow\nphysics. This Letter develops a model which incorporates kinetic effects in the\ngas, via the Boltzmann equation, and can predict experimentally-observed\nincreases in the maximum speed of wetting when (a) the liquid's viscosity is\nvaried, (b) the ambient gas pressure is reduced or (c) the meniscus is\nconfined.\n",
"title": "Kinetic Effects in Dynamic Wetting"
}
| null | null | null | null | true | null |
13464
| null |
Default
| null | null |
null |
{
"abstract": " The massive spread of digital misinformation has been identified as a major\nglobal risk and has been alleged to influence elections and threaten\ndemocracies. Communication, cognitive, social, and computer scientists are\nengaged in efforts to study the complex causes for the viral diffusion of\nmisinformation online and to develop solutions, while search and social media\nplatforms are beginning to deploy countermeasures. With few exceptions, these\nefforts have been mainly informed by anecdotal evidence rather than systematic\ndata. Here we analyze 14 million messages spreading 400 thousand articles on\nTwitter during and following the 2016 U.S. presidential campaign and election.\nWe find evidence that social bots played a disproportionate role in amplifying\nlow-credibility content. Accounts that actively spread articles from\nlow-credibility sources are significantly more likely to be bots. Automated\naccounts are particularly active in amplifying content in the very early\nspreading moments, before an article goes viral. Bots also target users with\nmany followers through replies and mentions. Humans are vulnerable to this\nmanipulation, retweeting bots who post links to low-credibility content.\nSuccessful low-credibility sources are heavily supported by social bots. These\nresults suggest that curbing social bots may be an effective strategy for\nmitigating the spread of online misinformation.\n",
"title": "The spread of low-credibility content by social bots"
}
| null | null | null | null | true | null |
13465
| null |
Default
| null | null |
null |
{
"abstract": " A vector field is called a Beltrami vector field, if $B\\times(\\nabla\\times\nB)=0$. In this paper we construct two unique Beltrami vector fields\n$\\mathfrak{I}$ and $\\mathfrak{Y}$, such that\n$\\nabla\\times\\mathfrak{I}=\\mathfrak{I}$,\n$\\nabla\\times\\mathfrak{Y}=\\mathfrak{Y}$, and such that both have an\norientation-preserving icosahedral symmetry. Both of them have an additional\nsymmetry with respect to a non-trivial automorphism of the number field\n$\\mathbb{Q}(\\,\\sqrt{5}\\,)$.\n",
"title": "Beltrami vector fields with an icosahedral symmetry"
}
| null | null | null | null | true | null |
13466
| null |
Default
| null | null |
null |
{
"abstract": " Couder and Fort discovered that droplets walking on a vibrating bath possess\ncertain features previously thought to be exclusive to quantum systems. These\nmillimetric droplets synchronize with their Faraday wavefield, creating a\nmacroscopic pilot-wave system. In this paper we exploit the fact that the waves\ngenerated are nearly monochromatic and propose a hydrodynamic model capable of\nquantitatively capturing the interaction between bouncing drops and a variable\ntopography. We show that our reduced model is able to reproduce some important\nexperiments involving the drop-topography interaction, such as non-specular\nreflection and single-slit diffraction.\n",
"title": "A model for Faraday pilot waves over variable topography"
}
| null | null | null | null | true | null |
13467
| null |
Default
| null | null |
null |
{
"abstract": " We investigate the languages recognized by well-structured transition systems\n(WSTS) with upward and downward compatibility. Our first result shows that,\nunder very mild assumptions, every two disjoint WSTS languages are regular\nseparable: There is a regular language containing one of them and being\ndisjoint from the other. As a consequence, if a language as well as its\ncomplement are both recognized by WSTS, then they are necessarily regular. In\nparticular, no subclass of WSTS languages beyond the regular languages is\nclosed under complement. Our second result shows that for Petri nets, the\ncomplexity of the backwards coverability algorithm yields a bound on the size\nof the regular separator. We complement it by a lower bound construction.\n",
"title": "Regular Separability of Well Structured Transition Systems"
}
| null | null | null | null | true | null |
13468
| null |
Default
| null | null |
null |
{
"abstract": " We introduce a new family of integrable stochastic processes, called\n\\textit{dynamical stochastic higher spin vertex models}, arising from fused\nrepresentations of Felder's elliptic quantum group $E_{\\tau, \\eta}\n(\\mathfrak{sl}_2)$. These models simultaneously generalize the stochastic\nhigher spin vertex models, studied by Corwin-Petrov and Borodin-Petrov, and are\ndynamical in the sense of Borodin's recent stochastic interaction round-a-face\nmodels.\nWe provide explicit contour integral identities for observables of these\nmodels (when run under specific types of initial data) that characterize the\ndistributions of their currents. Through asymptotic analysis of these\nidentities in a special case, we evaluate the scaling limit for the current of\na dynamical version of a discrete-time partial exclusion process. In\nparticular, we show that its scaling exponent is $1 / 4$ and that its one-point\nmarginal converges (in a sense of moments) to that of a non-trivial random\nvariable, which we determine explicitly.\n",
"title": "Dynamical Stochastic Higher Spin Vertex Models"
}
| null | null | null | null | true | null |
13469
| null |
Default
| null | null |
null |
{
"abstract": " A large number of applications such as querying sensor networks, and\nanalyzing protein-protein interaction (PPI) networks, rely on mining uncertain\ngraph and hypergraph databases. In this work we study the following problem:\ngiven an uncertain, weighted (hyper)graph, how can we efficiently find a\n(hyper)matching with high expected reward, and low risk?\nThis problem naturally arises in the context of several important\napplications, such as online dating, kidney exchanges, and team formation. We\nintroduce a novel formulation for finding matchings with maximum expected\nreward and bounded risk under a general model of uncertain weighted\n(hyper)graphs that we introduce in this work. Our model generalizes\nprobabilistic models used in prior work, and captures both continuous and\ndiscrete probability distributions, thus allowing to handle privacy related\napplications that inject appropriately distributed noise to (hyper)edge\nweights. Given that our optimization problem is NP-hard, we turn our attention\nto designing efficient approximation algorithms. For the case of uncertain\nweighted graphs, we provide a $\\frac{1}{3}$-approximation algorithm, and a\n$\\frac{1}{5}$-approximation algorithm with near optimal run time. For the case\nof uncertain weighted hypergraphs, we provide a\n$\\Omega(\\frac{1}{k})$-approximation algorithm, where $k$ is the rank of the\nhypergraph (i.e., any hyperedge includes at most $k$ nodes), that runs in\nalmost (modulo log factors) linear time.\nWe complement our theoretical results by testing our approximation algorithms\non a wide variety of synthetic experiments, where we observe in a controlled\nsetting interesting findings on the trade-off between reward, and risk. We also\nprovide an application of our formulation for providing recommendations of\nteams that are likely to collaborate, and have high impact.\n",
"title": "Risk-Averse Matchings over Uncertain Graph Databases"
}
| null | null | null | null | true | null |
13470
| null |
Default
| null | null |
null |
{
"abstract": " The context transformation and generalized context transformation methods, we\nintroduced recently, were able to reduce zero order entropy by exchanging\ndigrams, and as a consequence, they were removing mutual information between\nconsecutive symbols of the input message. These transformations were intended\nto be used as a preprocessor for zero-order entropy coding algorithms like\nArithmetic or Huffman coding, since we know, that especially Arithmetic coding\ncan achieve a compression rate almost of the size of Shannon's entropy.\nThis paper introduces a novel algorithm based on the concept of generalized\ncontext transformation, that allows transformation of words longer than simple\ndigrams. The higher order contexts are exploited using recursive form of a\ngeneralized context transformation. It is shown that the zero order entropy of\ntransformed data drops significantly, but on the other hand, the overhead given\nby a description of individual transformations increases and it has become a\nlimiting factor in a successful transformation of smaller files.\n",
"title": "Higher Order Context Transformations"
}
| null | null |
[
"Computer Science",
"Mathematics"
] | null | true | null |
13471
| null |
Validated
| null | null |
null |
{
"abstract": " In recent years numerous advanced malware, aka advanced persistent threats\n(APT) are allegedly developed by nation-states. The task of attributing an APT\nto a specific nation-state is extremely challenging for several reasons. Each\nnation-state has usually more than a single cyber unit that develops such\nadvanced malware, rendering traditional authorship attribution algorithms\nuseless. Furthermore, those APTs use state-of-the-art evasion techniques,\nmaking feature extraction challenging. Finally, the dataset of such available\nAPTs is extremely small.\nIn this paper we describe how deep neural networks (DNN) could be\nsuccessfully employed for nation-state APT attribution. We use sandbox reports\n(recording the behavior of the APT when run dynamically) as raw input for the\nneural network, allowing the DNN to learn high level feature abstractions of\nthe APTs itself. Using a test set of 1,000 Chinese and Russian developed APTs,\nwe achieved an accuracy rate of 94.6%.\n",
"title": "DeepAPT: Nation-State APT Attribution Using End-to-End Deep Neural Networks"
}
| null | null | null | null | true | null |
13472
| null |
Default
| null | null |
null |
{
"abstract": " Low-pressure gaseous TPCs are well suited detectors to correlate the\ndirections of nuclear recoils to the galactic Dark Matter (DM) halo. Indeed, in\naddition to providing a measure of the energy deposition due to the elastic\nscattering of a DM particle on a nucleus in the target gas, they allow for the\nreconstruction of the track of the recoiling nucleus. In order to exclude the\nbackground events originating from radioactive decays on the surfaces of the\ndetector materials within the drift volume, efforts are ongoing to precisely\nlocalize the track nuclear recoil in the drift volume along the axis\nperpendicular to the cathode plane. We report here the implementation of the\nmeasure of the signal induced on the cathode by the motion of the primary\nelectrons toward the anode in a MIMAC chamber. As a validation, we performed an\nindependent measurement of the drift velocity of the electrons in the\nconsidered gas mixture, correlating in time the cathode signal with the measure\nof the arrival times of the electrons on the anode.\n",
"title": "Cathode signal in a TPC directional detector: implementation and validation measuring the drift velocity"
}
| null | null | null | null | true | null |
13473
| null |
Default
| null | null |
null |
{
"abstract": " Due to the exponential complexity of the resources required by quantum state\ntomography (QST), people are interested in approaches towards identifying\nquantum states which require less effort and time. In this paper, we provide a\ntailored and efficient method for reconstructing mixed quantum states up to\n$12$ (or even more) qubits from an incomplete set of observables subject to\nnoises. Our method is applicable to any pure or nearly pure state $\\rho$, and\ncan be extended to many states of interest in quantum information processing,\nsuch as multi-particle entangled $W$ state, GHZ state and cluster states that\nare matrix product operators of low dimensions. The method applies the quantum\ndensity matrix constraints to a quantum compressive sensing optimization\nproblem, and exploits a modified Quantum Alternating Direction Multiplier\nMethod (Quantum-ADMM) to accelerate the convergence. Our algorithm takes $8,35$\nand $226$ seconds respectively to reconstruct superposition state density\nmatrices of $10,11,12$ qubits with acceptable fidelity, using less than $1 \\%$\nof measurements of expectation. To our knowledge it is the fastest realization\nthat people can achieve using a normal desktop. We further discuss applications\nof this method using experimental data of mixed states obtained in an ion trap\nexperiment of up to $8$ qubits.\n",
"title": "Fast Reconstruction of High-qubit Quantum States via Low Rate Measurements"
}
| null | null |
[
"Computer Science",
"Mathematics"
] | null | true | null |
13474
| null |
Validated
| null | null |
null |
{
"abstract": " Reconstructing network connectivity from the collective dynamics of a system\ntypically requires access to its complete continuous-time evolution although\nthese are often experimentally inaccessible. Here we propose a theory for\nrevealing physical connectivity of networked systems only from the event time\nseries their intrinsic collective dynamics generate. Representing the patterns\nof event timings in an event space spanned by inter-event and cross-event\nintervals, we reveal which other units directly influence the inter-event times\nof any given unit. For illustration, we linearize an event space mapping\nconstructed from the spiking patterns in model neural circuits to reveal the\npresence or absence of synapses between any pair of neurons as well as whether\nthe coupling acts in an inhibiting or activating (excitatory) manner. The\nproposed model-independent reconstruction theory is scalable to larger networks\nand may thus play an important role in the reconstruction of networks from\nbiology to social science and engineering.\n",
"title": "Inferring network connectivity from event timing patterns"
}
| null | null | null | null | true | null |
13475
| null |
Default
| null | null |
null |
{
"abstract": " Consider a set of agents that wish to estimate a vector of parameters of\ntheir mutual interest. For this estimation goal, agents can sense and\ncommunicate. When sensing, an agent measures (in additive gaussian noise)\nlinear combinations of the unknown vector of parameters. When communicating, an\nagent can broadcast information to a few other agents, by using the channels\nthat happen to be randomly at its disposal at the time.\nTo coordinate the agents towards their estimation goal, we propose a novel\nalgorithm called FADE (Fast and Asymptotically efficient Distributed\nEstimator), in which agents collaborate at discrete time-steps; at each\ntime-step, agents sense and communicate just once, while also updating their\nown estimate of the unknown vector of parameters.\nFADE enjoys five attractive features: first, it is an intuitive estimator,\nsimple to derive; second, it withstands dynamic networks, that is, networks\nwhose communication channels change randomly over time; third, it is strongly\nconsistent in that, as time-steps play out, each agent's local estimate\nconverges (almost surely) to the true vector of parameters; fourth, it is both\nasymptotically unbiased and efficient, which means that, across time, each\nagent's estimate becomes unbiased and the mean-square error (MSE) of each\nagent's estimate vanishes to zero at the same rate of the MSE of the optimal\nestimator at an almighty central node; fifth, and most importantly, when\ncompared with a state-of-art consensus+innovation (CI) algorithm, it yields\nestimates with outstandingly lower mean-square errors, for the same number of\ncommunications -- for example, in a sparsely connected network model with 50\nagents, we find through numerical simulations that the reduction can be\ndramatic, reaching several orders of magnitude.\n",
"title": "FADE: Fast and Asymptotically efficient Distributed Estimator for dynamic networks"
}
| null | null | null | null | true | null |
13476
| null |
Default
| null | null |
null |
{
"abstract": " During maintenance, software developers deal with a number of software change\nrequests. Each of those requests is generally written using natural language\ntexts, and it involves one or more domain related concepts. A developer needs\nto map those concepts to exact source code locations within the project in\norder to implement the requested change. This mapping generally starts with a\nsearch within the project that requires one or more suitable search terms.\nStudies suggest that the developers often perform poorly in coming up with good\nsearch terms for a change task. In this paper, we propose and evaluate a novel\nTextRank-based technique that automatically identifies and suggests search\nterms for a software change task by analyzing its task description. Experiments\nwith 349 change tasks from two subject systems and comparison with one of the\nlatest and closely related state-of-the-art approaches show that our technique\nis highly promising in terms of suggestion accuracy, mean average precision and\nrecall.\n",
"title": "TextRank Based Search Term Identification for Software Change Tasks"
}
| null | null | null | null | true | null |
13477
| null |
Default
| null | null |
null |
{
"abstract": " Upon employing the analysis in a new time domain, termed natural time, it has\nbeen recently demonstrated that a remarkable change of seismicity emerges\nbefore major mainshocks in California. What constitutes this change is that the\nfluctuations of the order parameter of seismicity exhibit a clearly detectable\nminimum. This is identified by using a natural time window sliding event by\nevent through the time series of the earthquakes in a wide area and comprising\na number of events that would occur on the average within a few months or so.\nHere, we suggest a method to estimate the epicentral area of an impending\nmainshock by an additional study of this minimum using an area window sliding\nthrough the wide area. We find that when this area window surrounds (or is\nadjacent to) the future epicentral area, the minimum of the order parameter\nfluctuations in this area appears at a date very close to the one at which the\nminimum is observed in the wide area. The method is applied here to major\nearthquakes that occurred in California during the recent decades including the\nlargest one, i.e., the 1992 Landers earthquake.\n",
"title": "Natural Time Analysis of Seismicity in California: The epicenter of an impending mainshock"
}
| null | null | null | null | true | null |
13478
| null |
Default
| null | null |
null |
{
"abstract": " Nanostructures have the immense potential to supplant the traditional\nmetallic structure as they show enhanced mechanical properties through strain\nhardening. In this paper, the effect of grain size on the hardening mechanism\nof Al-Cu nanostructure is elucidated by molecular dynamics simulation. Al-Cu\n(50-54% Cu by weight) nanostructure having an average grain size of 4.57 to\n7.26 nm are investigated for tensile simulation at different strain rate using\nembedded atom method (EAM) potential at a temperature of 50~500K. It is found\nthat the failure mechanism of the nanostructure is governed by the temperature,\ngrain size as well as strain rate effect. At the high temperature of 300-500K,\nthe failure strength of Al-Cu nanostructure increases with the decrease of\naverage grain size following Hall-Petch relation. Dislocation motions are\nhindered significantly when the grain size is decreased which play a vital role\non the hardening of the nanostructure. The failure is always found to initiate\nat a particular Al grain due to its weak link and propagates through grain\nboundary (GB) sliding, diffusion, dislocation nucleation and propagation. We\nalso visualize the dislocation density at different grain size to show how the\ndislocation affects the material properties at the nanoscale. These results\nwill further aid investigation on the deformation mechanism of nanostructure.\n",
"title": "Atomistic study of hardening mechanism in Al-Cu nanostructure"
}
| null | null | null | null | true | null |
13479
| null |
Default
| null | null |
null |
{
"abstract": " Future generation of gravitational wave detectors will have the sensitivity\nto detect gravitational wave events at redshifts far beyond any detectable\nelectromagnetic sources. We show that if the observed event rate is greater\nthan one event per year at redshifts z > 40, then the probability distribution\nof primordial density fluctuations must be significantly non-Gaussian or the\nevents originate from primordial black holes. The nature of the excess events\ncan be determined from the redshift distribution of the merger rate.\n",
"title": "Maximum redshift of gravitational wave merger events"
}
| null | null |
[
"Physics"
] | null | true | null |
13480
| null |
Validated
| null | null |
null |
{
"abstract": " The present work seeks to analyse the altmetric performance of Brazilian\npublications authored by researchers who are productivity scholarship holders\n(PQ) of the National Council of Scientific and Technological Development\n(CNPq). It was considered, within the scope of this research, the PQs in\nactivity in October, 2017 (n = 14.609). The scientific production registered on\nLattes was collected via GetLattesData and filtered by articles from academic\njournals published between 2016 and October 2017 that hold the Digital Object\nIdentifier (n = 99064). The online attention data are analysed according to\ntheir distribution by density and variation; language of the publication and\nfield of knowledge; and by average performance of the type of source that has\nprovided its altmetric values. The density evidences the long tail behavior of\nthe variable, with most part of the articles with altmetrics score = 0, while\nfew articles have a high index. The average of the online attention indicates a\nbetter performance of articles written in English and belonging to the Health\nand Biological Sciences field of knowledge. As for the sources, there was a\ngood performance from Mendeley, followed by Twitter and a low coverage from\nFacebook\n",
"title": "The altmetric performance of publications authored by Brazilian researchers: analysis of CNPq productivity scholarship holders"
}
| null | null | null | null | true | null |
13481
| null |
Default
| null | null |
null |
{
"abstract": " The impact of information dissemination on epidemic control essentially\naffects individual behaviors. Among the information-driven behaviors,\nvaccination is determined by the cost-related factors, and the correlation with\ninformation dissemination is not clear yet. To this end, we present a model to\nintegrate the information-epidemic spread process into an evolutionary\nvaccination game in multiplex networks, and explore how the spread of\ninformation on epidemic influences the vaccination behavior. We propose a\ntwo-layer coupled susceptible-alert-infected-susceptible (SAIS) model on a\nmultiplex network, where the strength coefficient is defined to characterize\nthe tendency and intensity of information dissemination. By means of the\nevolutionary game theory, we get the equilibrium vaccination level (the\nevolutionary stable strategy) for the vaccination game. After exploring the\ninfluence of the strength coefficient on the equilibrium vaccination level, we\nreach a counter-intuitive conclusion that more information transmission cannot\npromote vaccination. Specifically, when the vaccination cost is within a\ncertain range, increasing information dissemination even leads to a decline in\nthe equilibrium vaccination level. Moreover, we study the influence of the\nstrength coefficient on the infection density and social cost, and unveil the\nrole of information dissemination in controlling the epidemic with numerical\nsimulations.\n",
"title": "The Impact of Information Dissemination on Vaccinating Epidemics in Multiplex Networks"
}
| null | null | null | null | true | null |
13482
| null |
Default
| null | null |
null |
{
"abstract": " As artificial intelligence is increasingly affecting all parts of society and\nlife, there is growing recognition that human interpretability of machine\nlearning models is important. It is often argued that accuracy or other similar\ngeneralization performance metrics must be sacrificed in order to gain\ninterpretability. Such arguments, however, fail to acknowledge that the overall\ndecision-making system is composed of two entities: the learned model and a\nhuman who fuses together model outputs with his or her own information. As\nsuch, the relevant performance criteria should be for the entire system, not\njust for the machine learning component. In this work, we characterize the\nperformance of such two-node tandem data fusion systems using the theory of\ndistributed detection. In doing so, we work in the population setting and model\ninterpretable learned models as multi-level quantizers. We prove that under our\nabstraction, the overall system of a human with an interpretable classifier\noutperforms one with a black box classifier.\n",
"title": "Why Interpretability in Machine Learning? An Answer Using Distributed Detection and Data Fusion Theory"
}
| null | null | null | null | true | null |
13483
| null |
Default
| null | null |
null |
{
"abstract": " This paper presents sufficient conditions for the convergence of online\nestimation methods and the stability of adaptive control strategies for a class\nof history dependent, functional differential equations. The study is motivated\nby the increasing interest in estimation and control techniques for robotic\nsystems whose governing equations include history dependent nonlinearities. The\nfunctional differential equations in this paper are constructed using integral\noperators that depend on distributed parameters. As a consequence the resulting\nestimation and control equations are examples of distributed parameter systems\nwhose states and distributed parameters evolve in finite and infinite\ndimensional spaces, respectively. suWell-posedness, existence, and uniqueness\nare discussed for the class of fully actuated robotic systems with history\ndependent forces in their governing equation of motion. By deriving rates of\napproximation for the class of history dependent operators in this paper,\nsufficient conditions are derived that guarantee that finite dimensional\napproximations of the online estimation equations converge to the solution of\nthe infinite dimensional, distributed parameter system. The convergence and\nstability of a sliding mode adaptive control strategy for the history\ndependent, functional differential equations is established using Barbalat's\nlemma.\n",
"title": "Online Estimation and Adaptive Control for a Class of History Dependent Functional Differential Equations"
}
| null | null |
[
"Mathematics"
] | null | true | null |
13484
| null |
Validated
| null | null |
null |
{
"abstract": " We present spectroscopic observations of the C II $\\lambda$6578 permitted\nline for 83 lines of sight in 76 planetary nebulae at high spectral resolution,\nmost of them obtained with the Manchester Echelle Spectrograph on the 2.1\\,m\ntelescope at the Observatorio Astronómico Nacional on the Sierra San Pedro\nMártir. We study the kinematics of the C II $\\lambda$6578 permitted line with\nrespect to other permitted and collisionally-excited lines. Statistically, we\nfind that the kinematics of the C II $\\lambda$6578 line are not those expected\nif this line arises from the recombination of C$^{2+}$ ions or the fluorescence\nof C$^+$ ions in ionization equilibrium in a chemically-homogeneous nebular\nplasma, but instead its kinematics are those appropriate for a volume more\ninternal than expected. The planetary nebulae in this sample have well-defined\nmorphology and are restricted to a limited range in H$\\alpha$ line widths (no\nlarge values) compared to their counterparts in the Milky Way bulge, both of\nwhich could be interpreted as the result of young nebular shells, an inference\nthat is also supported by nebular modeling. Concerning the long-standing\ndiscrepancy between chemical abundances inferred from permitted and\ncollisionally-excited emission lines in photoionized nebulae, our results imply\nthat multiple plasma components occur commonly in planetary nebulae.\n",
"title": "The Kinematics of the Permitted C II $λ$ 6578 Line in a Large Sample of Planetary Nebulae"
}
| null | null | null | null | true | null |
13485
| null |
Default
| null | null |
null |
{
"abstract": " Metasurfaces are promising tools towards novel designs for flat optics\napplications. As such their quality and tolerance to fabrication imperfections\nneed to be evaluated with specific tools. However, most such tools rely on the\ngeometrical optics approximation and are not straightforwardly applicable to\nmetasurfaces. In this Letter, we introduce and evaluate, for metasurfaces,\nparameters such as the intercept factor and the slope error usually defined for\nsolar concentrators in the realm of ray-optics. After proposing definitions\nvalid in physical optics, we put forward an approach to calculate them. As\nexamples, we design three different concentrators based on three specific unit\ncells and assess them numerically. The concept allows for the comparison of the\nefficiency of the metasurfaces, their sensitivities to fabrication\nimperfections and will be critical for practical systems.\n",
"title": "From parabolic-trough to metasurface-concentrator"
}
| null | null | null | null | true | null |
13486
| null |
Default
| null | null |
null |
{
"abstract": " We prove that the open Gromov-Witten invariants on K3 surfaces satisfy the\nKontsevich-Soibelman wall-crossing formula. One one hand, this gives a\ngeometric interpretation of the slab functions in Gross-Siebert program. On the\nother hands, the open Gromov-Witten invariants coincide with the weighted\ncounting of tropical discs. This is an analog of the corresponding theorem on\ntoric varieties \\cite{M2}\\cite{NS} but on compact Calabi-Yau surfaces.\n",
"title": "Correspondence Theorem between Holomorphic Discs and Tropical Discs on K3 Surfaces"
}
| null | null | null | null | true | null |
13487
| null |
Default
| null | null |
null |
{
"abstract": " This paper reviews the main estimation and prediction results derived in the\ncontext of functional time series, when Hilbert and Banach spaces are\nconsidered, specially, in the context of autoregressive processes of order one\n(ARH(1) and ARB(1) processes, for H and B being a Hilbert and Banach space,\nrespectively). Particularly, we pay attention to the estimation and prediction\nresults, and statistical tests, derived in both parametric and non-parametric\nframeworks. A comparative study between different ARH(1) prediction approaches\nis developed in the simulation study undertaken.\n",
"title": "A review and comparative study on functional time series techniques"
}
| null | null | null | null | true | null |
13488
| null |
Default
| null | null |
null |
{
"abstract": " When rough grains in standard packing conditions are discharged from a silo,\na conical depression with a single slope is formed at the surface. We observed\nthat the increase of the volume fraction generates a more complex depression\ncharacterized by two angles of discharge: a lower angle close to the one\nmeasured for standard packing and a considerably larger upper angle. The change\nin slope appears at the boundary between a densely packed stagnant region at\nthe periphery and the central flowing channel formed over the aperture. Since\nthe material in the latter zone is always fluidized, the flow rate is\nunaffected by the initial packing of the bed. On the other hand, the contrast\nbetween both angles is markedly smaller when smooth particles of the same size\nand density are used, which reveals that high volume fraction and friction must\ncombine to produce the observed geometry. Our results show that the surface\nprofile helps to identify by simple visual inspection the packing conditions of\na granular bed, and this can be useful to prevent undesirable collapses during\nsilo discharge in industry.\n",
"title": "Surface depression with double-angle geometry during the discharge of close-packed grains from a silo"
}
| null | null | null | null | true | null |
13489
| null |
Default
| null | null |
null |
{
"abstract": " Non-stationary domains, where unforeseen changes happen, present a challenge\nfor agents to find an optimal policy for a sequential decision making problem.\nThis work investigates a solution to this problem that combines Markov Decision\nProcesses (MDP) and Reinforcement Learning (RL) with Answer Set Programming\n(ASP) in a method we call ASP(RL). In this method, Answer Set Programming is\nused to find the possible trajectories of an MDP, from where Reinforcement\nLearning is applied to learn the optimal policy of the problem. Results show\nthat ASP(RL) is capable of efficiently finding the optimal solution of an MDP\nrepresenting non-stationary domains.\n",
"title": "Answer Set Programming for Non-Stationary Markov Decision Processes"
}
| null | null | null | null | true | null |
13490
| null |
Default
| null | null |
null |
{
"abstract": " An important theorem in classical complexity theory is that LOGLOGSPACE=REG,\ni.e. that languages decidable with double-logarithmic space bound are regular.\nWe consider a transfinite analogue of this theorem. To this end, we introduce\ndeterministic ordinal automata (DOAs), show that they satisfy many of the basic\nstatements of the theory of deterministic finite automata and regular\nlanguages. We then consider languages decidable by an ordinal Turing machine\n(OTM), introduced by P. Koepke in 2005 and show that if the working space of an\nOTM is of strictly smaller cardinality than the input length for all\nsufficiently long inputs, the language so decided is also decidable by a DOA.\n",
"title": "Space-Bounded OTMs and REG$^{\\infty}$"
}
| null | null |
[
"Mathematics"
] | null | true | null |
13491
| null |
Validated
| null | null |
null |
{
"abstract": " We present the first polynomial-time approximation schemes, i.e., (1 +\n{\\epsilon})-approximation algorithm for any constant {\\epsilon} > 0, for the\nminimum three-edge connected spanning subgraph problem and the minimum\nthree-vertex connected spanning subgraph problem in undirected planar graphs.\nBoth the approximation schemes run in linear time.\n",
"title": "Linear-time approximation schemes for planar minimum three-edge connected and three-vertex connected spanning subgraphs"
}
| null | null | null | null | true | null |
13492
| null |
Default
| null | null |
null |
{
"abstract": " The emergence of smart Wi-Fi APs (Access Point), which are equipped with huge\nstorage space, opens a new research area on how to utilize these resources at\nthe edge network to improve users' quality of experience (QoE) (e.g., a short\nstartup delay and smooth playback). One important research interest in this\narea is content prefetching, which predicts and accurately fetches contents\nahead of users' requests to shift the traffic away during peak periods.\nHowever, in practice, the different video watching patterns among users, and\nthe varying network connection status lead to the time-varying server load,\nwhich eventually makes the content prefetching problem challenging. To\nunderstand this challenge, this paper first performs a large-scale measurement\nstudy on users' AP connection and TV series watching patterns using\nreal-traces. Then, based on the obtained insights, we formulate the content\nprefetching problem as a Markov Decision Process (MDP). The objective is to\nstrike a balance between the increased prefetching&storage cost incurred by\nincorrect prediction and the reduced content download delay because of\nsuccessful prediction. A learning-based approach is proposed to solve this\nproblem and another three algorithms are adopted as baselines. In particular,\nfirst, we investigate the performance lower bound by using a random algorithm,\nand the upper bound by using an ideal offline approach. Then, we present a\nheuristic algorithm as another baseline. Finally, we design a reinforcement\nlearning algorithm that is more practical to work in the online manner. Through\nextensive trace-based experiments, we demonstrate the performance gain of our\ndesign. Remarkably, our learning-based algorithm achieves a better precision\nand hit ratio (e.g., 80%) with about 70% (resp. 50%) cost saving compared to\nthe random (resp. heuristic) algorithm.\n",
"title": "Towards Wi-Fi AP-Assisted Content Prefetching for On-Demand TV Series: A Reinforcement Learning Approach"
}
| null | null | null | null | true | null |
13493
| null |
Default
| null | null |
null |
{
"abstract": " The Abella interactive theorem prover has proven to be an effective vehicle\nfor reasoning about relational specifications. However, the system has a\nlimitation that arises from the fact that it is based on a simply typed logic:\nformalizations that are identical except in the respect that they apply to\ndifferent types have to be repeated at each type. We develop an approach that\novercomes this limitation while preserving the logical underpinnings of the\nsystem. In this approach object constructors, formulas and other relevant\nlogical notions are allowed to be parameterized by types, with the\ninterpretation that they stand for the (infinite) collection of corresponding\nconstructs that are obtained by instantiating the type parameters. The proof\nstructures that we consider for formulas that are schematized in this fashion\nare limited to ones whose type instances are valid proofs in the simply typed\nlogic. We develop schematic proof rules that ensure this property, a task that\nis complicated by the fact that type information influences the notion of\nunification that plays a key role in the logic. Our ideas, which have been\nimplemented in an updated version of the system, accommodate schematic\npolymorphism both in the core logic of Abella and in the executable\nspecification logic that it embeds.\n",
"title": "Schematic Polymorphism in the Abella Proof Assistant"
}
| null | null | null | null | true | null |
13494
| null |
Default
| null | null |
null |
{
"abstract": " An arithmetic matroid is weakly multiplicative if the multiplicity of at\nleast one of its bases is equal to the product of the multiplicities of its\nelements. We show that if such an arithmetic matroid can be represented by an\ninteger matrix, then this matrix is uniquely determined. This implies that the\ninteger cohomology ring of a centred toric arrangement whose arithmetic matroid\nis weakly multiplicative is determined by its poset of layers. This partially\nanswers a question asked by Callegaro-Delucchi.\n",
"title": "Representations of weakly multiplicative arithmetic matroids are unique"
}
| null | null | null | null | true | null |
13495
| null |
Default
| null | null |
null |
{
"abstract": " Multi-label submodular Markov Random Fields (MRFs) have been shown to be\nsolvable using max-flow based on an encoding of the labels proposed by\nIshikawa, in which each variable $X_i$ is represented by $\\ell$ nodes (where\n$\\ell$ is the number of labels) arranged in a column. However, this method in\ngeneral requires $2\\,\\ell^2$ edges for each pair of neighbouring variables.\nThis makes it inapplicable to realistic problems with many variables and\nlabels, due to excessive memory requirement. In this paper, we introduce a\nvariant of the max-flow algorithm that requires much less storage.\nConsequently, our algorithm makes it possible to optimally solve multi-label\nsubmodular problems involving large numbers of variables and labels on a\nstandard computer.\n",
"title": "Memory Efficient Max Flow for Multi-label Submodular MRFs"
}
| null | null | null | null | true | null |
13496
| null |
Default
| null | null |
null |
{
"abstract": " We present a strategy to obtain explicit equations for the modular double\ncovers associated respectively to both a split and a non-split Cartan subgroup\nof $\\text{GL}_2(\\mathbb F_{p})$ with $p$ prime. Then we apply it successfully\nto the level $13$ case.\n",
"title": "Double Covers of Cartan Modular Curves"
}
| null | null | null | null | true | null |
13497
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we present a simple analysis of {\\bf fast rates} with {\\it\nhigh probability} of {\\bf empirical minimization} for {\\it stochastic composite\noptimization} over a finite-dimensional bounded convex set with exponential\nconcave loss functions and an arbitrary convex regularization. To the best of\nour knowledge, this result is the first of its kind. As a byproduct, we can\ndirectly obtain the fast rate with {\\it high probability} for exponential\nconcave empirical risk minimization with and without any convex regularization,\nwhich not only extends existing results of empirical risk minimization but also\nprovides a unified framework for analyzing exponential concave empirical risk\nminimization with and without {\\it any} convex regularization. Our proof is\nvery simple only exploiting the covering number of a finite-dimensional bounded\nset and a concentration inequality of random vectors.\n",
"title": "A Simple Analysis for Exp-concave Empirical Minimization with Arbitrary Convex Regularizer"
}
| null | null | null | null | true | null |
13498
| null |
Default
| null | null |
null |
{
"abstract": " By using numerical and analytical methods, we describe the generation of\nfine-scale lateral electromagnetic waves, called surface plasmon-polaritons\n(SPPs), on atomically thick, metamaterial conducting sheets in two spatial\ndimensions (2D). Our computations capture the two-scale character of the total\nfield and reveal how each edge of the sheet acts as a source of an SPP that may\ndominate the diffracted field. We use the finite element method to numerically\nimplement a variational formulation for a weak discontinuity of the tangential\nmagnetic field across a hypersurface. An adaptive, local mesh refinement\nstrategy based on a posteriori error estimators is applied to resolve the\npronounced two-scale character of wave propagation and radiation over the\nmetamaterial sheet. We demonstrate by numerical examples how a singular\ngeometry, e.g., sheets with sharp edges, and sharp spatial changes in the\nassociated surface conductivity may significantly influence surface plasmons in\nnanophotonics.\n",
"title": "Generation of surface plasmon-polaritons by edge effects"
}
| null | null | null | null | true | null |
13499
| null |
Default
| null | null |
null |
{
"abstract": " As deep neural networks become more complex and input datasets grow larger,\nit can take days or even weeks to train a deep neural network to the desired\naccuracy. Therefore, distributed Deep Learning at a massive scale is a critical\ncapability, since it offers the potential to reduce the training time from\nweeks to hours. In this paper, we present a software-hardware co-optimized\ndistributed Deep Learning system that can achieve near-linear scaling up to\nhundreds of GPUs. The core algorithm is a multi-ring communication pattern that\nprovides a good tradeoff between latency and bandwidth and adapts to a variety\nof system configurations. The communication algorithm is implemented as a\nlibrary for easy use. This library has been integrated into Tensorflow, Caffe,\nand Torch. We train Resnet-101 on Imagenet 22K with 64 IBM Power8 S822LC\nservers (256 GPUs) in about 7 hours to an accuracy of 33.8 % validation\naccuracy. Microsoft's ADAM and Google's DistBelief results did not reach 30 %\nvalidation accuracy for Imagenet 22K. Compared to Facebook AI Research's recent\npaper on 256 GPU training, we use a different communication algorithm, and our\ncombined software and hardware system offers better communication overhead for\nResnet-50. A PowerAI DDL enabled version of Torch completed 90 epochs of\ntraining on Resnet 50 for 1K classes in 50 minutes using 64 IBM Power8 S822LC\nservers (256 GPUs).\n",
"title": "PowerAI DDL"
}
| null | null | null | null | true | null |
13500
| null |
Default
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.