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": " Resonance energy transfer (RET) is an inherently anisotropic process. Even\nthe simplest, well-known Förster theory, based on the transition\ndipole-dipole coupling, implicitly incorporates the anisotropic character of\nRET. In this theoretical work, we study possible signatures of the fundamental\nanisotropic character of RET in hybrid nanomaterials composed of a\nsemiconductor nanoparticle (NP) decorated with molecular dyes. In particular,\nby means of a realistic kinetic model, we show that the analysis of the dye\nphotoluminescence difference for orthogonal input polarizations reveals the\nanisotropic character of the dye-NP RET which arises from the intrinsic\nanisotropy of the NP lattice. In a prototypical core/shell wurtzite CdSe/ZnS NP\nfunctionalized with cyanine dyes (Cy3B), this difference is predicted to be as\nlarge as 75\\% and it is strongly dependent in amplitude and sign on the dye-NP\ndistance. We account for all the possible RET processes within the system,\ntogether with competing decay pathways in the separate segments. In addition,\nwe show that the anisotropic signature of RET is persistent up to a large\nnumber of dyes per NP.\n",
"title": "Predicting signatures of anisotropic resonance energy transfer in dye-functionalized nanoparticles"
}
| null | null | null | null | true | null |
6201
| null |
Default
| null | null |
null |
{
"abstract": " Single individual haplotyping is an NP-hard problem that emerges when\nattempting to reconstruct an organism's inherited genetic variations using data\ntypically generated by high-throughput DNA sequencing platforms. Genomes of\ndiploid organisms, including humans, are organized into homologous pairs of\nchromosomes that differ from each other in a relatively small number of variant\npositions. Haplotypes are ordered sequences of the nucleotides in the variant\npositions of the chromosomes in a homologous pair; for diploids, haplotypes\nassociated with a pair of chromosomes may be conveniently represented by means\nof complementary binary sequences. In this paper, we consider a binary matrix\nfactorization formulation of the single individual haplotyping problem and\nefficiently solve it by means of alternating minimization. We analyze the\nconvergence properties of the alternating minimization algorithm and establish\ntheoretical bounds for the achievable haplotype reconstruction error. The\nproposed technique is shown to outperform existing methods when applied to\nsynthetic as well as real-world Fosmid-based HapMap NA12878 datasets.\n",
"title": "Matrix Completion and Performance Guarantees for Single Individual Haplotyping"
}
| null | null |
[
"Statistics",
"Quantitative Biology"
] | null | true | null |
6202
| null |
Validated
| null | null |
null |
{
"abstract": " Cellular or dendritic microstructures that result as a function of additive\nmanufacturing solidification conditions in a Ni-based melt pool are simulated\nin the present work using three-dimensional phase-field simulations. A\nmacroscopic thermal model is used to obtain the temperature gradient $G$ and\nthe solidification velocity $V$ which are provided as inputs to the phase-field\nmodel. We extract the cell spacings, cell core compositions, and cell tip as\nwell as mushy zone temperatures from the simulated microstructures as a\nfunction of $V$. Cell spacings are compared with different scaling laws that\ncorrelate to the solidification conditions and approximated by $G^{-m}V^{-n}$.\nCell core compositions are compared with the analytical solutions of a dendrite\ngrowth theory and found to be in good agreement. Through analysis of the mushy\nzone, we extract a characteristic bridging plane, where the primary $\\gamma$\nphase coalesces across the intercellular liquid channels at a $\\gamma$ fraction\nbetween 0.6 and 0.7. The temperature and the $\\gamma$ fraction in this plane\nare found to decrease with increasing $V$. The simulated microstructural\nfeatures are significant as they can be used as inputs for the simulation of\nsubsequent heat treatment processes.\n",
"title": "Simulation and analysis of $γ$-Ni cellular growth during laser powder deposition of Ni-based superalloys"
}
| null | null | null | null | true | null |
6203
| null |
Default
| null | null |
null |
{
"abstract": " A quantum system of particles can exist in a localized phase, exhibiting\nergodicity breaking and maintaining forever a local memory of its initial\nconditions. We generalize this concept to a system of extended objects, such as\nstrings and membranes, arguing that such a system can also exhibit localization\nin the presence of sufficiently strong disorder (randomness) in the\nHamiltonian. We show that localization of large extended objects can be mapped\nto a lower-dimensional many-body localization problem. For example, motion of a\nstring involves propagation of point-like signals down its length to keep the\ndifferent segments in causal contact. For sufficiently strong disorder, all\nsuch internal modes will exhibit many-body localization, resulting in the\nlocalization of the entire string. The eigenstates of the system can then be\nconstructed perturbatively through a convergent 'string locator expansion.' We\npropose a type of out-of-time-order string correlator as a diagnostic of such a\nstring localized phase. Localization of other higher-dimensional objects, such\nas membranes, can also be studied through a hierarchical construction by\nmapping onto localization of lower-dimensional objects. Our arguments are\n'asymptotic' ($i.e.$ valid up to rare regions) but they extend the notion of\nlocalization (and localization protected order) to a host of settings where\nsuch ideas previously did not apply. These include high-dimensional\nferromagnets with domain wall excitations, three-dimensional topological phases\nwith loop-like excitations, and three-dimensional type-II superconductors with\nflux line excitations. In type-II superconductors, localization of flux lines\ncould stabilize superconductivity at energy densities where a normal state\nwould arise in thermal equilibrium.\n",
"title": "Localization of Extended Quantum Objects"
}
| null | null | null | null | true | null |
6204
| null |
Default
| null | null |
null |
{
"abstract": " Projective Reed-Muller codes were introduced by Lachaud, in 1988 and their\ndimension and minimum distance were determined by Serre and S{\\o}rensen in\n1991. In coding theory one is also interested in the higher Hamming weights, to\nstudy the code performance. Yet, not many values of the higher Hamming weights\nare known for these codes, not even the second lowest weight (also known as\nnext-to-minimal weight) is completely determined. In this paper we determine\nall the values of the next-to-minimal weight for the binary projective\nReed-Muller codes, which we show to be equal to the next-to-minimal weight of\nReed-Muller codes in most, but not all, cases.\n",
"title": "The next-to-minimal weights of binary projective Reed-Muller codes"
}
| null | null | null | null | true | null |
6205
| null |
Default
| null | null |
null |
{
"abstract": " We present an algorithm for classification tasks on big data. Experiments\nconducted as part of this study indicate that the algorithm can be as accurate\nas ensemble methods such as random forests or gradient boosted trees. Unlike\nensemble methods, the models produced by the algorithm can be easily\ninterpreted. The algorithm is based on a divide and conquer strategy and\nconsists of two steps. The first step consists of using a decision tree to\nsegment the large dataset. By construction, decision trees attempt to create\nhomogeneous class distributions in their leaf nodes. However, non-homogeneous\nleaf nodes are usually produced. The second step of the algorithm consists of\nusing a suitable classifier to determine the class labels for the\nnon-homogeneous leaf nodes. The decision tree segment provides a coarse segment\nprofile while the leaf level classifier can provide information about the\nattributes that affect the label within a segment.\n",
"title": "Big Data Classification Using Augmented Decision Trees"
}
| null | null | null | null | true | null |
6206
| null |
Default
| null | null |
null |
{
"abstract": " We study a set of uniquely determined tilting and cotilting modules for an\nalgebra with positive dominant dimension, with the property that they are\ngenerated or cogenerated (and usually both) by projective-injectives. These\nmodules have various interesting properties, for example that their\nendomorphism algebras always have global dimension at most that of the original\nalgebra. We characterise d-Auslander-Gorenstein algebras and d-Auslander\nalgebras via the property that the relevant tilting and cotilting modules\ncoincide. By the Morita-Tachikawa correspondence, any algebra of dominant\ndimension at least 2 may be expressed (essentially uniquely) as the\nendomorphism algebra of a generator-cogenerator for another algebra, and we\nalso study our special tilting and cotilting modules from this point of view,\nvia the theory of recollements and intermediate extension functors.\n",
"title": "Special tilting modules for algebras with positive dominant dimension"
}
| null | null | null | null | true | null |
6207
| null |
Default
| null | null |
null |
{
"abstract": " Corruptive behaviour in politics limits economic growth, embezzles public\nfunds, and promotes socio-economic inequality in modern democracies. We analyse\nwell-documented political corruption scandals in Brazil over the past 27 years,\nfocusing on the dynamical structure of networks where two individuals are\nconnected if they were involved in the same scandal. Our research reveals that\ncorruption runs in small groups that rarely comprise more than eight people, in\nnetworks that have hubs and a modular structure that encompasses more than one\ncorruption scandal. We observe abrupt changes in the size of the largest\nconnected component and in the degree distribution, which are due to the\ncoalescence of different modules when new scandals come to light or when\ngovernments change. We show further that the dynamical structure of political\ncorruption networks can be used for successfully predicting partners in future\nscandals. We discuss the important role of network science in detecting and\nmitigating political corruption.\n",
"title": "The dynamical structure of political corruption networks"
}
| null | null | null | null | true | null |
6208
| null |
Default
| null | null |
null |
{
"abstract": " Knowing a biomolecule's structure is inherently linked to and a prerequisite\nfor any detailed understanding of its function. Significant effort has gone\ninto developing technologies for structural characterization. These\ntechnologies do not directly provide 3D structures; instead they typically\nyield noisy and erroneous distance information between specific entities such\nas atoms or residues, which have to be translated into consistent 3D models.\nHere we present an approach for this translation process based on\nmaxent-stress optimization. Our new approach extends the original graph drawing\nmethod for the new application's specifics by introducing additional\nconstraints and confidence values as well as algorithmic components. Extensive\nexperiments demonstrate that our approach infers structural models (i. e.,\nsensible 3D coordinates for the molecule's atoms) that correspond well to the\ndistance information, can handle noisy and error-prone data, and is\nconsiderably faster than established tools. Our results promise to allow domain\nscientists nearly-interactive structural modeling based on distance\nconstraints.\n",
"title": "Maxent-Stress Optimization of 3D Biomolecular Models"
}
| null | null |
[
"Computer Science",
"Physics"
] | null | true | null |
6209
| null |
Validated
| null | null |
null |
{
"abstract": " We study the quantum synchronization between a pair of two-level systems\ninside two coupled cavities. By using a digital-analog decomposition of the\nmaster equation that rules the system dynamics, we show that this approach\nleads to quantum synchronization between both two-level systems. Moreover, we\ncan identify in this digital-analog block decomposition the fundamental\nelements of a quantum machine learning protocol, in which the agent and the\nenvironment (learning units) interact through a mediating system, namely, the\nregister. If we can additionally equip this algorithm with a classical feedback\nmechanism, which consists of projective measurements in the register,\nreinitialization of the register state and local conditional operations on the\nagent and environment subspace, a powerful and flexible quantum machine\nlearning protocol emerges. Indeed, numerical simulations show that this\nprotocol enhances the synchronization process, even when every subsystem\nexperience different loss/decoherence mechanisms, and give us the flexibility\nto choose the synchronization state. Finally, we propose an implementation\nbased on current technologies in superconducting circuits.\n",
"title": "Enhanced Quantum Synchronization via Quantum Machine Learning"
}
| null | null | null | null | true | null |
6210
| null |
Default
| null | null |
null |
{
"abstract": " We propose and experimentally demonstrate a technique for coupling phonons\nout of an optomechanical crystal cavity. By designing a perturbation that\nbreaks a symmetry in the elastic structure, we selectively induce phonon\nleakage without affecting the optical properties. It is shown experimentally\nvia cryogenic measurements that the proposed cavity perturbation causes loss of\nphonons into mechanical waves on the surface of silicon, while leaving photon\nlifetimes unaffected. This demonstrates that phonon leakage can be engineered\nin on-chip optomechanical systems. We experimentally observe large fluctuations\nin leakage rates that we attribute to fabrication disorder and verify this\nusing simulations. Our technique opens the way to engineering more complex\non-chip phonon networks utilizing guided mechanical waves to connect quantum\nsystems.\n",
"title": "Engineering phonon leakage in nanomechanical resonators"
}
| null | null | null | null | true | null |
6211
| null |
Default
| null | null |
null |
{
"abstract": " Since the discovery of the Meissner effect the superconductor to normal (S-N)\nphase transition in the presence of a magnetic field is understood to be a\nfirst order phase transformation that is reversible under ideal conditions and\nobeys the laws of thermodynamics. The reverse (N-S) transition is the Meissner\neffect. This implies in particular that the kinetic energy of the supercurrent\nis not dissipated as Joule heat in the process where the superconductor becomes\nnormal and the supercurrent stops. In this paper we analyze the entropy\ngeneration and the momentum transfer between the supercurrent and the body in\nthe S-N transition and the N-S transition as described by the conventional\ntheory of superconductivity. We find that it is impossible to explain the\ntransition in a way that is consistent with the laws of thermodynamics unless\nthe momentum transfer between the supercurrent and the body occurs with zero\nentropy generation, for which the conventional theory of superconductivity\nprovides no mechanism. Instead, we point out that the alternative theory of\nhole superconductivity does not encounter such difficulties.\n",
"title": "Entropy generation and momentum transfer in the superconductor-normal and normal-superconductor phase transformations and the consistency of the conventional theory of superconductivity"
}
| null | null | null | null | true | null |
6212
| null |
Default
| null | null |
null |
{
"abstract": " We prove a local Faber-Krahn inequality for solutions $u$ to the Dirichlet\nproblem for $\\Delta + V$ on an arbitrary domain $\\Omega$ in $\\mathbb{R}^n$.\nSuppose a solution $u$ assumes a global maximum at some point $x_0 \\in \\Omega$\nand $u(x_0)>0$. Let $T(x_0)$ be the smallest time at which a Brownian motion,\nstarted at $x_0$, has exited the domain $\\Omega$ with probability $\\ge 1/2$.\nFor nice (e.g., convex) domains, $T(x_0) \\asymp d(x_0,\\partial\\Omega)^2$ but we\nmake no assumption on the geometry of the domain. Our main result is that there\nexists a ball $B$ of radius $\\asymp T(x_0)^{1/2}$ such that $$ \\| V\n\\|_{L^{\\frac{n}{2}, 1}(\\Omega \\cap B)} \\ge c_n > 0, $$ provided that $n \\ge 3$.\nIn the case $n = 2$, the above estimate fails and we obtain a substitute\nresult. The Laplacian may be replaced by a uniformly elliptic operator in\ndivergence form. This result both unifies and strenghtens a series of earlier\nresults.\n",
"title": "A Local Faber-Krahn inequality and Applications to Schrödinger's Equation"
}
| null | null | null | null | true | null |
6213
| null |
Default
| null | null |
null |
{
"abstract": " These notes were written as supplementary material for a five-hour lecture\nseries presented at the Centre de Recerca Mathemàtica at the Universitat\nAutònoma de Barcelona from the 13th to the 17th of March 2017. The intention\nof these notes is to give a brief overview of some key topics in the area of\n$C^*$-algebras associated to étale groupoids. The scope has been deliberately\ncontained to the case of étale groupoids with the intention that much of the\nrepresentation-theoretic technology and measure-theoretic analysis required to\nhandle general groupoids can be suppressed in this simpler setting.\nA published version of these notes will appear in the volume tentatively\ntitled \"Operator algebras and dynamics: groupoids, crossed products and Rokhlin\ndimension\" by Gabor Szabo, Dana P. Williams and myself, and edited by Francesc\nPerera, in the series \"Advanced Courses in Mathematics. CRM Barcelona.\" The\npagination of this arXiv version is not identical to Birkhäuser's style, but\nI have tried to make it close. The theorem numbering should be correct. I'm\ngrateful to the CRM and Birkhäuser for allowing me to post a version on\narXiv.\n",
"title": "Étale groupoids and their $C^*$-algebras"
}
| null | null | null | null | true | null |
6214
| null |
Default
| null | null |
null |
{
"abstract": " In a recent work, Bindini and De Pascale have introduced a regularization of\n$N$-particle symmetric probabilities which preserves their one-particle\nmarginals. In this short note, we extend their construction to mixed quantum\nfermionic states. This enables us to prove the convergence of the Levy-Lieb\nfunctional in Density Functional Theory , to the corresponding multi-marginal\noptimal transport in the semi-classical limit. Our result holds for mixed\nstates of any particle number $N$, with or without spin.\n",
"title": "Semi-classical limit of the Levy-Lieb functional in Density Functional Theory"
}
| null | null | null | null | true | null |
6215
| null |
Default
| null | null |
null |
{
"abstract": " Let $ \\alpha: \\mathcal{C} \\to \\mathcal{D}$ be a symmetric monoidal functor\nfrom a stable presentable symmetric monoidal $\\infty$-category $\\mathcal{C} $\ncompactly generated by the tensorunit to a stable presentable symmetric\nmonoidal $\\infty$-category $ \\mathcal{D} $ with compact tensorunit. Let $\\beta:\n\\mathcal{D} \\to \\mathcal{C}$ be a right adjoint of $\\alpha$ and $ \\mathrm{X}:\n\\mathcal{B} \\to \\mathcal{D} $ a symmetric monoidal functor starting at a small\nrigid symmetric monoidal $\\infty$-category $ \\mathcal{B}$. We construct a\nsymmetric monoidal equivalence between modules in the $\\infty$-category of\nfunctors $ \\mathcal{B} \\to \\mathcal{C} $ over the $ \\mathrm{E}_\\infty$-algebra\n$\\beta \\circ \\mathrm{X} $ and the full subcategory of $\\mathcal{D}$ compactly\ngenerated by the essential image of $\\mathrm{X}$. Especially for every motivic\n$ \\mathrm{E}_\\infty$-ring spectrum $\\mathrm{A}$ we obtain a symmetric monoidal\nequivalence between the $\\infty$-category of cellular motivic\n$\\mathrm{A}$-module spectra and modules in the $\\infty$-category of functors\n$\\mathrm{QS}$ to spectra over some $ \\mathrm{E}_\\infty$-algebra, where\n$\\mathrm{QS}$ denotes the 0th space of the sphere spectrum.\n",
"title": "A characterization of cellular motivic spectra"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6216
| null |
Validated
| null | null |
null |
{
"abstract": " Summarizes recent work on the wakefields and impedances of flat, metallic\nplates with small corrugations\n",
"title": "The Impedance of Flat Metallic Plates with Small Corrugations"
}
| null | null | null | null | true | null |
6217
| null |
Default
| null | null |
null |
{
"abstract": " Community detection in graphs is the problem of finding groups of vertices\nwhich are more densely connected than they are to the rest of the graph. This\nproblem has a long history, but it is undergoing a resurgence of interest due\nto the need to analyze social and biological networks. While there are many\nways to formalize it, one of the most popular is as an inference problem, where\nthere is a \"ground truth\" community structure built into the graph somehow. The\ntask is then to recover the ground truth knowing only the graph.\nRecently it was discovered, first heuristically in physics and then\nrigorously in probability and computer science, that this problem has a phase\ntransition at which it suddenly becomes impossible. Namely, if the graph is too\nsparse, or the probabilistic process that generates it is too noisy, then no\nalgorithm can find a partition that is correlated with the planted one---or\neven tell if there are communities, i.e., distinguish the graph from a purely\nrandom one with high probability. Above this information-theoretic threshold,\nthere is a second threshold beyond which polynomial-time algorithms are known\nto succeed; in between, there is a regime in which community detection is\npossible, but conjectured to require exponential time.\nFor computer scientists, this field offers a wealth of new ideas and open\nquestions, with connections to probability and combinatorics, message-passing\nalgorithms, and random matrix theory. Perhaps more importantly, it provides a\nwindow into the cultures of statistical physics and statistical inference, and\nhow those cultures think about distributions of instances, landscapes of\nsolutions, and hardness.\n",
"title": "The Computer Science and Physics of Community Detection: Landscapes, Phase Transitions, and Hardness"
}
| null | null |
[
"Computer Science",
"Physics",
"Mathematics"
] | null | true | null |
6218
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper, we consider a dataset comprising press releases about health\nresearch from different universities in the UK along with a corresponding set\nof news articles. First, we do an exploratory analysis to understand how the\nbasic information published in the scientific journals get exaggerated as they\nare reported in these press releases or news articles. This initial analysis\nshows that some news agencies exaggerate almost 60\\% of the articles they\npublish in the health domain; more than 50\\% of the press releases from certain\nuniversities are exaggerated; articles in topics like lifestyle and childhood\nare heavily exaggerated. Motivated by the above observation we set the central\nobjective of this paper to investigate how exaggerated news spreads over an\nonline social network like Twitter. The LIWC analysis points to a remarkable\nobservation these late tweets are essentially laden in words from opinion and\nrealize categories which indicates that, given sufficient time, the wisdom of\nthe crowd is actually able to tell apart the exaggerated news. As a second step\nwe study the characteristics of the users who never or rarely post exaggerated\nnews content and compare them with those who post exaggerated news content more\nfrequently. We observe that the latter class of users have less retweets or\nmentions per tweet, have significantly more number of followers, use more slang\nwords, less hyperbolic words and less word contractions. We also observe that\nthe LIWC categories like bio, health, body and negative emotion are more\npronounced in the tweets posted by the users in the latter class. As a final\nstep we use these observations as features and automatically classify the two\ngroups achieving an F1 score of 0.83.\n",
"title": "Characterizing the spread of exaggerated news content over social media"
}
| null | null | null | null | true | null |
6219
| null |
Default
| null | null |
null |
{
"abstract": " This paper investigates the effects of finite flat porous extensions to\nsemi-infinite impermeable flat plates in an attempt to control trailing-edge\nnoise through bio-inspired adaptations. Specifically the problem of sound\ngenerated by a gust convecting in uniform mean steady flow scattering off the\ntrailing edge and permeable-impermeable junction is considered. This setup\nsupposes that any realistic trailing-edge adaptation to a blade would be\nsufficiently small so that the turbulent boundary layer encapsulates both the\nporous edge and the permeable-impermeable junction, and therefore the\ninteraction of acoustics generated at these two discontinuous boundaries is\nimportant. The acoustic problem is tackled analytically through use of the\nWiener-Hopf method. A two-dimensional matrix Wiener-Hopf problem arises due to\nthe two interaction points (the trailing edge and the permeable-impermeable\njunction). This paper discusses a new iterative method for solving this matrix\nWiener-Hopf equation which extends to further two-dimensional problems in\nparticular those involving analytic terms that exponentially grow in the upper\nor lower half planes. This method is an extension of the commonly used \"pole\nremoval\" technique and avoids the needs for full matrix factorisation.\nConvergence of this iterative method to an exact solution is shown to be\nparticularly fast when terms neglected in the second step are formally smaller\nthan all other terms retained. The final acoustic solution highlights the\neffects of the permeable-impermeable junction on the generated noise, in\nparticular how this junction affects the far-field noise generated by\nhigh-frequency gusts by creating an interference to typical trailing-edge\nscattering. This effect results in partially porous plates predicting a lower\nnoise reduction than fully porous plates when compared to fully impermeable\nplates.\n",
"title": "Aerodynamic noise from rigid trailing edges with finite porous extensions"
}
| null | null | null | null | true | null |
6220
| null |
Default
| null | null |
null |
{
"abstract": " It is well known that many optimization methods, including SGD, SAGA, and\nAccelerated SGD for over-parameterized models, do not scale linearly in the\nparallel setting. In this paper, we present a new version of block coordinate\ndescent that solves this issue for a number of methods. The core idea is to\nmake the sampling of coordinate blocks on each parallel unit independent of the\nothers. Surprisingly, we prove that the optimal number of blocks to be updated\nby each of $n$ units in every iteration is equal to $m/n$, where $m$ is the\ntotal number of blocks. As an illustration, this means that when $n=100$\nparallel units are used, $99\\%$ of work is a waste of time. We demonstrate that\nwith $m/n$ blocks used by each unit the iteration complexity often remains the\nsame. Among other applications which we mention, this fact can be exploited in\nthe setting of distributed optimization to break the communication bottleneck.\nOur claims are justified by numerical experiments which demonstrate almost a\nperfect match with our theory on a number of datasets.\n",
"title": "99% of Parallel Optimization is Inevitably a Waste of Time"
}
| null | null | null | null | true | null |
6221
| null |
Default
| null | null |
null |
{
"abstract": " The control of spins and spin to charge conversion in organics requires\nunderstanding the molecular spin-orbit coupling (SOC), and a means to tune its\nstrength. However, quantifying SOC strengths indirectly through spin relaxation\neffects has proven diffi- cult due to competing relaxation mechanisms. Here we\npresent a systematic study of the g-tensor shift in molecular semiconductors\nand link it directly to the SOC strength in a series of high mobility molecular\nsemiconductors with strong potential for future devices. The results\ndemonstrate a rich variability of the molecular g-shifts with the effective\nSOC, depending on subtle aspects of molecular composition and structure. We\ncorrelate the above g -shifts to spin-lattice relaxation times over four orders\nof magnitude, from 200 {\\mu}s to 0.15 {\\mu}s, for isolated molecules in\nsolution and relate our findings for isolated molecules in solution to the spin\nrelaxation mechanisms that are likely to be relevant in solid state systems.\n",
"title": "Tuning the effective spin-orbit coupling in molecular semiconductors"
}
| null | null | null | null | true | null |
6222
| null |
Default
| null | null |
null |
{
"abstract": " Recent research has shown the potential utility of Deep Gaussian Processes.\nThese deep structures are probability distributions, designed through\nhierarchical construction, which are conditionally Gaussian. In this paper, the\ncurrent published body of work is placed in a common framework and, through\nrecursion, several classes of deep Gaussian processes are defined. The\nresulting samples generated from a deep Gaussian process have a Markovian\nstructure with respect to the depth parameter, and the effective depth of the\nresulting process is interpreted in terms of the ergodicity, or non-ergodicity,\nof the resulting Markov chain. For the classes of deep Gaussian processes\nintroduced, we provide results concerning their ergodicity and hence their\neffective depth. We also demonstrate how these processes may be used for\ninference; in particular we show how a Metropolis-within-Gibbs construction\nacross the levels of the hierarchy can be used to derive sampling tools which\nare robust to the level of resolution used to represent the functions on a\ncomputer. For illustration, we consider the effect of ergodicity in some simple\nnumerical examples.\n",
"title": "How Deep Are Deep Gaussian Processes?"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
6223
| null |
Validated
| null | null |
null |
{
"abstract": " While single measurement vector (SMV) models have been widely studied in\nsignal processing, there is a surging interest in addressing the multiple\nmeasurement vectors (MMV) problem. In the MMV setting, more than one\nmeasurement vector is available and the multiple signals to be recovered share\nsome commonalities such as a common support. Applications in which MMV is a\nnaturally occurring phenomenon include online streaming, medical imaging, and\nvideo recovery. This work presents a stochastic iterative algorithm for the\nsupport recovery of jointly sparse corrupted MMV. We present a variant of the\nSparse Randomized Kaczmarz algorithm for corrupted MMV and compare our proposed\nmethod with an existing Kaczmarz type algorithm for MMV problems. We also\nshowcase the usefulness of our approach in the online (streaming) setting and\nprovide empirical evidence that suggests the robustness of the proposed method\nto the distribution of the corruption and the number of corruptions occurring.\n",
"title": "Sparse Randomized Kaczmarz for Support Recovery of Jointly Sparse Corrupted Multiple Measurement Vectors"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6224
| null |
Validated
| null | null |
null |
{
"abstract": " Dynamical materials that capable of responding to optical stimuli have always\nbeen pursued for designing novel photonic devices and functionalities, of which\nthe response speed and amplitude as well as integration adaptability and energy\neffectiveness are especially critical. Here we show ultrafast pulse generation\nby exploiting the ultrafast and sensitive nonlinear dynamical processes in\ntunably solution-processed colloidal epsilon-near-zero (ENZ) transparent\nconducting oxide (TCO) nanocrystals (NCs), of which the potential respond\nresponse speed is >2 THz and modulation depth is ~23% pumped at ~0.7 mJ/cm2,\nbenefiting from the highly confined geometry in addition to the ENZ enhancement\neffect. These ENZ NCs may offer a scalable and printable material solution for\ndynamic photonic and optoelectronic devices.\n",
"title": "Exploiting ITO colloidal nanocrystals for ultrafast pulse generation"
}
| null | null |
[
"Physics"
] | null | true | null |
6225
| null |
Validated
| null | null |
null |
{
"abstract": " Categorization is necessary for many decision making tasks. However, the\ncategorization process may interfere the decision making result and the law of\ntotal probability can be violated in some situations. To predict the\ninterference effect of categorization, some model based on quantum probability\nhas been proposed. In this paper, a new quantum dynamic belief (QDB) model is\nproposed. Considering the precise decision may not be made during the process,\nthe concept of uncertainty is introduced in our model to simulate real human\nthinking process. Then the interference effect categorization can be predicted\nby handling the uncertain information. The proposed model is applied to a\ncategorization decision-making experiment to explain the interference effect of\ncategorization. Compared with other models, our model is relatively more\nsuccinct and the result shows the correctness and effectiveness of our model.\n",
"title": "A quantum dynamic belief model to explain the interference effects of categorization on decision making"
}
| null | null | null | null | true | null |
6226
| null |
Default
| null | null |
null |
{
"abstract": " Millimeter wave communications rely on narrow-beam transmissions to cope with\nthe strong signal attenuation at these frequencies, thus demanding precise beam\nalignment between transmitter and receiver. The communication overhead incurred\nto achieve beam alignment may become a severe impairment in mobile networks.\nThis paper addresses the problem of optimizing beam alignment acquisition, with\nthe goal of maximizing throughput. Specifically, the algorithm jointly\ndetermines the portion of time devoted to beam alignment acquisition, as well\nas, within this portion of time, the optimal beam search parameters, using the\nframework of Markov decision processes. It is proved that a bisection search\nalgorithm is optimal, and that it outperforms exhaustive and iterative search\nalgorithms proposed in the literature. The duration of the beam alignment phase\nis optimized so as to maximize the overall throughput. The numerical results\nshow that the throughput, optimized with respect to the duration of the beam\nalignment phase, achievable under the exhaustive algorithm is 88.3% lower than\nthat achievable under the bisection algorithm. Similarly, the throughput\nachievable by the iterative search algorithm for a division factor of 4 and 8\nis, respectively, 12.8% and 36.4% lower than that achievable by the bisection\nalgorithm.\n",
"title": "Throughput Optimal Beam Alignment in Millimeter Wave Networks"
}
| null | null | null | null | true | null |
6227
| null |
Default
| null | null |
null |
{
"abstract": " This workshop invites researchers and practitioners to participate in\nexploring behavioral change support intelligent transportation applications. We\nwelcome submissions that explore intelligent transportation systems (ITS),\nwhich interact with travelers in order to persuade them or nudge them towards\nsustainable transportation behaviors and decisions. Emerging opportunities\nincluding the use of data and information generated by ITS and users' mobile\ndevices in order to render personalized, contextualized and timely transport\nbehavioral change interventions are in our focus. We invite submissions and\nideas from domains of ITS including, but not limited to, multi-modal journey\nplanners, advanced traveler information systems and in-vehicle systems. The\nexpected outcome will be a deeper understanding of the challenges and future\nresearch directions with respect to behavioral change support through ITS.\n",
"title": "Behavioural Change Support Intelligent Transportation Applications"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6228
| null |
Validated
| null | null |
null |
{
"abstract": " Motivated by the description of Nurowski's conformal structure for maximally\nsymmetric homogeneous examples of bracket-generating rank 2 distributions in\ndimension 5, aka $(2,3,5)$-distributions, we consider a rank $3$ Pfaffian\nsystem in dimension 5 with $SU(2)$ symmetry. We find the conditions for which\nthis Pfaffian system has the maximal symmetry group (in the real case this is\nthe split real form of $G_2$), and give the associated Nurowski's conformal\nclasses. We also present a $SU(2)$ gauge-theoretic interpretation of the\nresults obtained.\n",
"title": "SU(2) Pfaffian systems and gauge theory"
}
| null | null | null | null | true | null |
6229
| null |
Default
| null | null |
null |
{
"abstract": " We study the effect of electron correlations on a system consisting of a\nsingle-level quantum dot with local Coulomb interaction attached to two\nsuperconducting leads. We use the single-impurity Anderson model with BCS\nsuperconducting baths to study the interplay between the proximity induced\nelectron pairing and the local Coulomb interaction. We show how to solve the\nmodel using the continuous-time hybridization-expansion quantum Monte Carlo\nmethod. The results obtained for experimentally relevant parameters are\ncompared with results of self-consistent second order perturbation theory as\nwell as with the numerical renormalization group method.\n",
"title": "Correlation effects in superconducting quantum dot systems"
}
| null | null | null | null | true | null |
6230
| null |
Default
| null | null |
null |
{
"abstract": " We propose a method for learning Markov network structures for continuous\ndata without invoking any assumptions about the distribution of the variables.\nThe method makes use of previous work on a non-parametric estimator for mutual\ninformation which is used to create a non-parametric test for multivariate\nconditional independence. This independence test is then combined with an\nefficient constraint-based algorithm for learning the graph structure. The\nperformance of the method is evaluated on several synthetic data sets and it is\nshown to learn considerably more accurate structures than competing methods\nwhen the dependencies between the variables involve non-linearities.\n",
"title": "Learning non-parametric Markov networks with mutual information"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6231
| null |
Validated
| null | null |
null |
{
"abstract": " In this work, we design a machine learning based method, online adaptive\nprimal support vector regression (SVR), to model the implied volatility surface\n(IVS). The algorithm proposed is the first derivation and implementation of an\nonline primal kernel SVR. It features enhancements that allow efficient online\nadaptive learning by embedding the idea of local fitness and budget maintenance\nto dynamically update support vectors upon pattern drifts. For algorithm\nacceleration, we implement its most computationally intensive parts in a Field\nProgrammable Gate Arrays hardware, where a 132x speedup over CPU is achieved\nduring online prediction. Using intraday tick data from the E-mini S&P 500\noptions market, we show that the Gaussian kernel outperforms the linear kernel\nin regulating the size of support vectors, and that our empirical IVS algorithm\nbeats two competing online methods with regards to model complexity and\nregression errors (the mean absolute percentage error of our algorithm is up to\n13%). Best results are obtained at the center of the IVS grid due to its larger\nnumber of adjacent support vectors than the edges of the grid. Sensitivity\nanalysis is also presented to demonstrate how hyper parameters affect the error\nrates and model complexity.\n",
"title": "Online Adaptive Machine Learning Based Algorithm for Implied Volatility Surface Modeling"
}
| null | null | null | null | true | null |
6232
| null |
Default
| null | null |
null |
{
"abstract": " This study proposes a control strategy for the efficient semi active\nsuspension systems utilizing a novel hybrid PID-fuzzy logic control scheme .In\nthe control architecture, we employ the Chaotic Fruit Fly Algorithm for PID\ntuning since it can avoid local minima by chaotic search. A novel linguistic\nrule based fuzzy logic controller is developed to aid the PID.A quarter car\nmodel with a non-linear spring system is used to test the performance of the\nproposed control approach. A road terrain is chosen where the comfort and\nhandling parameters are tested specifically in the regions of abrupt changes.\nThe results suggest that the suspension systems controlled by the hybrid\nstrategy has the potential to offer more comfort and handling by reducing the\npeak acceleration and suspension distortion by 83.3 % and 28.57% respectively\nwhen compared to the active suspension systems. Also, compared to the\nperformance of similar suspension control strategies optimized by stochastic\nalgorithms such as Genetic Algorithms (GA), Particle Swarm Optimization (PSO)\nand Bacterial Foraging Optimization (BFO), reductions in peak acceleration and\nsuspension distortion are found to be 25%, 32.3%, 54.6% and 23.35 %, 22.5%, 5.4\n% respectively.The details of the solution methodology have been presented in\nthe paper.\n",
"title": "Hybrid control strategy for a semi active suspension system using fuzzy logic and bio-inspired chaotic fruit fly algorithm"
}
| null | null | null | null | true | null |
6233
| null |
Default
| null | null |
null |
{
"abstract": " We demonstrate that, for a range of state-of-the-art machine learning\nalgorithms, the differences in generalisation performance obtained using\ndefault parameter settings and using parameters tuned via cross-validation can\nbe similar in magnitude to the differences in performance observed between\nstate-of-the-art and uncompetitive learning systems. This means that fair and\nrigorous evaluation of new learning algorithms requires performance comparison\nagainst benchmark methods with best-practice model selection procedures, rather\nthan using default parameter settings. We investigate the sensitivity of three\nkey machine learning algorithms (support vector machine, random forest and\nrotation forest) to their default parameter settings, and provide guidance on\ndetermining sensible default parameter values for implementations of these\nalgorithms. We also conduct an experimental comparison of these three\nalgorithms on 121 classification problems and find that, perhaps surprisingly,\nrotation forest is significantly more accurate on average than both random\nforest and a support vector machine.\n",
"title": "On the Use of Default Parameter Settings in the Empirical Evaluation of Classification Algorithms"
}
| null | null | null | null | true | null |
6234
| null |
Default
| null | null |
null |
{
"abstract": " In further study of the application of crossed-product functors to the\nBaum-Connes Conjecture, Buss, Echterhoff, and Willett introduced various other\nproperties that crossed-product functors may have. Here we introduce and study\nanalogues of these properties for coaction functors, making sure that the\nproperties are preserved when the coaction functors are composed with the full\ncrossed product to make a crossed-product functor. The new properties for\ncoaction functors studied here are functoriality for generalized homomorphisms\nand the correspondence property. We particularly study the connections with the\nideal property. The study of functoriality for generalized homomorphisms\nrequires a detailed development of the Fischer construction of maximalization\nof coactions with regard to possibly degenerate homomorphisms into multiplier\nalgebras. We verify that all \"KLQ\" functors arising from large ideals of the\nFourier-Stieltjes algebra $B(G)$ have all the properties we study, and at the\nopposite extreme we give an example of a coaction functor having none of the\nproperties.\n",
"title": "Coaction functors, II"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6235
| null |
Validated
| null | null |
null |
{
"abstract": " We describe a complete list of Casimirs for 2D Euler hydrodynamics on a\nsurface without boundary: we define generalized enstrophies which, along with\ncirculations, form a complete set of invariants for coadjoint orbits of\narea-preserving diffeomorphisms on a surface. We also outline a possible\nextension of main notions to the boundary case and formulate several open\nquestions in that setting.\n",
"title": "Classification of Casimirs in 2D hydrodynamics"
}
| null | null | null | null | true | null |
6236
| null |
Default
| null | null |
null |
{
"abstract": " This paper presents a novel design of a crawler robot which is capable of\ntransforming its chassis from an Omni crawler mode to a large-sized wheel mode\nusing a novel mechanism. The transformation occurs without any additional\nactuators. Interestingly the robot can transform into a large diameter and\nsmall width wheel which enhances its maneuverability like small turning radius\nand fast/efficient locomotion. This paper contributes on improving the\nlocomotion mode of previously developed hybrid compliant omnicrawler robot\nCObRaSO. In addition to legged and tracked mechanism, CObRaSO can now display\nlarge wheel mode which contributes to its locomotion capabilities. Mechanical\ndesign of the robot has been explained in a detailed manner in this paper and\nalso the transforming experiment and torque analysis has been shown clearly\n",
"title": "Novel Compliant omnicrawler-wheel transforming module"
}
| null | null | null | null | true | null |
6237
| null |
Default
| null | null |
null |
{
"abstract": " A bifurcation is a qualitative change in a family of solutions to an equation\nproduced by varying parameters. In contrast to the local bifurcations of\ndynamical systems that are often related to a change in the number or stability\nof equilibria, bifurcations of boundary value problems are global in nature and\nmay not be related to any obvious change in dynamical behaviour. Catastrophe\ntheory is a well-developed framework which studies the bifurcations of critical\npoints of functions. In this paper we study the bifurcations of solutions of\nboundary-value problems for symplectic maps, using the language of\n(finite-dimensional) singularity theory. We associate certain such problems\nwith a geometric picture involving the intersection of Lagrangian submanifolds,\nand hence with the critical points of a suitable generating function. Within\nthis framework, we then study the effect of three special cases: (i) some\ncommon boundary conditions, such as Dirichlet boundary conditions for\nsecond-order systems, restrict the possible types of bifurcations (for example,\nin generic planar systems only the A-series beginning with folds and cusps can\noccur); (ii) integrable systems, such as planar Hamiltonian systems, can\nexhibit a novel periodic pitchfork bifurcation; and (iii) systems with\nHamiltonian symmetries or reversing symmetries can exhibit restricted\nbifurcations associated with the symmetry. This approach offers an alternative\nto the analysis of critical points in function spaces, typically used in the\nstudy of bifurcation of variational problems, and opens the way to the\ndetection of more exotic bifurcations than the simple folds and cusps that are\noften found in examples.\n",
"title": "Bifurcation of solutions to Hamiltonian boundary value problems"
}
| null | null | null | null | true | null |
6238
| null |
Default
| null | null |
null |
{
"abstract": " Manual annotations of temporal bounds for object interactions (i.e. start and\nend times) are typical training input to recognition, localization and\ndetection algorithms. For three publicly available egocentric datasets, we\nuncover inconsistencies in ground truth temporal bounds within and across\nannotators and datasets. We systematically assess the robustness of\nstate-of-the-art approaches to changes in labeled temporal bounds, for object\ninteraction recognition. As boundaries are trespassed, a drop of up to 10% is\nobserved for both Improved Dense Trajectories and Two-Stream Convolutional\nNeural Network.\nWe demonstrate that such disagreement stems from a limited understanding of\nthe distinct phases of an action, and propose annotating based on the Rubicon\nBoundaries, inspired by a similarly named cognitive model, for consistent\ntemporal bounds of object interactions. Evaluated on a public dataset, we\nreport a 4% increase in overall accuracy, and an increase in accuracy for 55%\nof classes when Rubicon Boundaries are used for temporal annotations.\n",
"title": "Trespassing the Boundaries: Labeling Temporal Bounds for Object Interactions in Egocentric Video"
}
| null | null | null | null | true | null |
6239
| null |
Default
| null | null |
null |
{
"abstract": " One-dimensional (1D) electron systems in the presence of Coulomb interaction\nare described by Luttinger liquid theory. The strength of Coulomb interaction\nin the Luttinger liquid, as parameterized by the Luttinger parameter K, is in\ngeneral difficult to measure. This is because K is usually hidden in powerlaw\ndependencies of observables as a function of temperature or applied bias. We\npropose a dynamical way to measure K on the basis of an electronic\ntime-of-flight experiment. We argue that the helical Luttinger liquid at the\nedge of a 2D topological insulator constitutes a preeminently suited\nrealization of a 1D system to test our proposal. This is based on the\nrobustness of helical liquids against elastic backscattering in the presence of\ntime reversal symmetry.\n",
"title": "Dynamical transport measurement of the Luttinger parameter in helical edges states of 2D topological insulators"
}
| null | null | null | null | true | null |
6240
| null |
Default
| null | null |
null |
{
"abstract": " Our experiment shows that the thermal emission of phonon can be controlled by\nmagnetic resonance (MR) mode in a metasurface (MTS). Through changing the\nstructural parameter of metasurface, the MR wavelength can be tuned to the\nphonon resonance wavelength. This introduces a strong coupling between phonon\nand MR, which results in an anticrossing phonon-plasmons mode. In the process,\nwe can manipulate the polarization and angular radiation of thermal emission of\nphonon. Such metasurface provides a new kind of thermal emission structures for\nvarious thermal management applications.\n",
"title": "Controlling thermal emission of phonon by magnetic metasurfaces"
}
| null | null |
[
"Physics"
] | null | true | null |
6241
| null |
Validated
| null | null |
null |
{
"abstract": " This paper addresses distributed average tracking of physical second-order\nagents with heterogeneous nonlinear dynamics, where there is no constraint on\ninput signals. The nonlinear terms in agents' dynamics are heterogeneous,\nsatisfying a Lipschitz-like condition that will be defined later and is more\ngeneral than the Lipschitz condition. In the proposed algorithm, a control\ninput and a filter are designed for each agent. Each agent's filter has two\noutputs and the idea is that the first output estimates the average of the\ninput signals and the second output estimates the average of the input\nvelocities asymptotically. In parallel, each agent's position and velocity are\ndriven to track, respectively, the first and the second outputs. Having\nheterogeneous nonlinear terms in agents' dynamics necessitates designing the\nfilters for agents. Since the nonlinear terms in agents' dynamics can be\nunbounded and the input signals are arbitrary, novel state-dependent\ntime-varying gains are employed in agents' filters and control inputs to\novercome these unboundedness effects. Finally the results are improved to\nachieve the distributed average tracking for a group of double-integrator\nagents, where there is no constraint on input signals and the filter is not\nrequired anymore. Numerical simulations are also presented to illustrate the\ntheoretical results.\n",
"title": "Distributed Average Tracking of Heterogeneous Physical Second-order Agents With No Input Signals Constraint"
}
| null | null | null | null | true | null |
6242
| null |
Default
| null | null |
null |
{
"abstract": " We study the two-dimensional massless Dirac equation for a potential that is\nallowed to depend on the energy and on one of the spatial variables. After\ndetermining a modified orthogonality relation and norm for such systems, we\npresent an application involving an energy-dependent version of the hyperbolic\nScarf potential. We construct closed-form bound state solutions of the\nassociated Dirac equation.\n",
"title": "Bound states of the two-dimensional Dirac equation for an energy-dependent hyperbolic Scarf potential"
}
| null | null | null | null | true | null |
6243
| null |
Default
| null | null |
null |
{
"abstract": " Humans take advantage of real world symmetries for various tasks, yet\ncapturing their superb symmetry perception mechanism with a computational model\nremains elusive. Motivated by a new study demonstrating the extremely high\ninter-person accuracy of human perceived symmetries in the wild, we have\nconstructed the first deep-learning neural network for reflection and rotation\nsymmetry detection (Sym-NET), trained on photos from MS-COCO (Microsoft-Common\nObject in COntext) dataset with nearly 11K consistent symmetry-labels from more\nthan 400 human observers. We employ novel methods to convert discrete human\nlabels into symmetry heatmaps, capture symmetry densely in an image and\nquantitatively evaluate Sym-NET against multiple existing computer vision\nalgorithms. On CVPR 2013 symmetry competition testsets and unseen MS-COCO\nphotos, Sym-NET significantly outperforms all other competitors. Beyond\nmathematically well-defined symmetries on a plane, Sym-NET demonstrates\nabilities to identify viewpoint-varied 3D symmetries, partially occluded\nsymmetrical objects, and symmetries at a semantic level.\n",
"title": "Beyond Planar Symmetry: Modeling human perception of reflection and rotation symmetries in the wild"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6244
| null |
Validated
| null | null |
null |
{
"abstract": " We consider the fractional Hartree equation in the $L^2$-supercritical case,\nand we find a sharp threshold of the scattering versus blow-up dichotomy for\nradial data: If $\nM[u_{0}]^{\\frac{s-s_c}{s_c}}E[u_{0}<M[Q]^{\\frac{s-s_c}{s_c}}E[Q]$ and\n$M[u_{0}]^{\\frac{s-s_c}{s_c}}\\|u_{0}\\|^2_{\\dot H^s}<M[Q]^{\\frac{s-s_c}{s_c}}\\|\nQ\\|^2_{\\dot H^s}$, then the solution $u(t)$ is globally well-posed and\nscatters; if $\nM[u_{0}]^{\\frac{s-s_c}{s_c}}E[u_{0}]<M[Q]^{\\frac{s-s_c}{s_c}}E[Q]$ and\n$M[u_{0}]^{\\frac{s-s_c}{s_c}}\\|u_{0}\\|^2_{\\dot H^s}>M[Q]^{\\frac{s-s_c}{s_c}}\\|\nQ\\|^2_{\\dot H^s}$, the solution $u(t)$ blows up in finite time. This condition\nis sharp in the sense that the solitary wave solution $e^{it}Q(x)$ is global\nbut not scattering, which satisfies the equality in the above conditions. Here,\n$Q$ is the ground-state solution for the fractional Hartree equation.\n",
"title": "Sharp Threshold of Blow-up and Scattering for the fractional Hartree equation"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6245
| null |
Validated
| null | null |
null |
{
"abstract": " This work studies the problem of stochastic dynamic filtering and state\npropagation with complex beliefs. The main contribution is GP-SUM, a filtering\nalgorithm tailored to dynamic systems and observation models expressed as\nGaussian Processes (GP), and to states represented as a weighted sum of\nGaussians. The key attribute of GP-SUM is that it does not rely on\nlinearizations of the dynamic or observation models, or on unimodal Gaussian\napproximations of the belief, hence enables tracking complex state\ndistributions. The algorithm can be seen as a combination of a sampling-based\nfilter with a probabilistic Bayes filter. On the one hand, GP-SUM operates by\nsampling the state distribution and propagating each sample through the dynamic\nsystem and observation models. On the other hand, it achieves effective\nsampling and accurate probabilistic propagation by relying on the GP form of\nthe system, and the sum-of-Gaussian form of the belief. We show that GP-SUM\noutperforms several GP-Bayes and Particle Filters on a standard benchmark. We\nalso demonstrate its use in a pushing task, predicting with experimental\naccuracy the naturally occurring non-Gaussian distributions.\n",
"title": "GP-SUM. Gaussian Processes Filtering of non-Gaussian Beliefs"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6246
| null |
Validated
| null | null |
null |
{
"abstract": " We consider an optimal stopping problem where a constraint is placed on the\ndistribution of the stopping time. Reformulating the problem in terms of\nso-called measure-valued martingales allows us to transform the marginal\nconstraint into an initial condition and view the problem as a stochastic\ncontrol problem; we establish the corresponding dynamic programming principle.\n",
"title": "A Dynamic Programming Principle for Distribution-Constrained Optimal Stopping"
}
| null | null | null | null | true | null |
6247
| null |
Default
| null | null |
null |
{
"abstract": " We give a survey of a generalization of Quillen-Sullivan rational homotopy\ntheory which gives spectral algebra models of unstable v_n-periodic homotopy\ntypes. In addition to describing and contextualizing our original approach, we\nsketch two other recent approaches which are of a more conceptual nature, due\nto Arone-Ching and Heuts. In the process, we also survey many relevant concepts\nwhich arise in the study of spectral algebra over operads, including\ntopological André-Quillen cohomology, Koszul duality, and Goodwillie\ncalculus.\n",
"title": "Spectral algebra models of unstable v_n-periodic homotopy theory"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6248
| null |
Validated
| null | null |
null |
{
"abstract": " The FitzHugh-Nagumo equation provides a simple mathematical model of cardiac\ntissue as an excitable medium hosting spiral wave vortices. Here we present\nextensive numerical simulations studying long-term dynamics of knotted vortex\nstring solutions for all torus knots up to crossing number 11. We demonstrate\nthat FitzHugh-Nagumo evolution preserves the knot topology for all the examples\npresented, thereby providing a novel field theory approach to the study of\nknots. Furthermore, the evolution yields a well-defined minimal length for each\nknot that is comparable to the ropelength of ideal knots. We highlight the role\nof the medium boundary in stabilizing the length of the knot and discuss the\nimplications beyond torus knots. By applying Moffatt's test we are able to show\nthat there is not a unique attractor within a given knot topology.\n",
"title": "The length of excitable knots"
}
| null | null | null | null | true | null |
6249
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we give a conditional lower bound of $n^{\\Omega(k)}$ on\nrunning time for the classic k-median and k-means clustering objectives (where\nn is the size of the input), even in low-dimensional Euclidean space of\ndimension four, assuming the Exponential Time Hypothesis (ETH). We also\nconsider k-median (and k-means) with penalties where each point need not be\nassigned to a center, in which case it must pay a penalty, and extend our lower\nbound to at least three-dimensional Euclidean space.\nThis stands in stark contrast to many other geometric problems such as the\ntraveling salesman problem, or computing an independent set of unit spheres.\nWhile these problems benefit from the so-called (limited) blessing of\ndimensionality, as they can be solved in time $n^{O(k^{1-1/d})}$ or\n$2^{n^{1-1/d}}$ in d dimensions, our work shows that widely-used clustering\nobjectives have a lower bound of $n^{\\Omega(k)}$, even in dimension four.\nWe complete the picture by considering the two-dimensional case: we show that\nthere is no algorithm that solves the penalized version in time less than\n$n^{o(\\sqrt{k})}$, and provide a matching upper bound of $n^{O(\\sqrt{k})}$.\nThe main tool we use to establish these lower bounds is the placement of\npoints on the moment curve, which takes its inspiration from constructions of\npoint sets yielding Delaunay complexes of high complexity.\n",
"title": "The Bane of Low-Dimensionality Clustering"
}
| null | null | null | null | true | null |
6250
| null |
Default
| null | null |
null |
{
"abstract": " We present the properties and advantages of a new magneto-optical trap (MOT)\nwhere blue-detuned light drives `type-II' transitions that have dark ground\nstates. Using $^{87}$Rb, we reach a radiation-pressure-limited density\nexceeding $10^{11}$cm$^{-3}$ and a temperature below 30$\\mu$K. The phase-space\ndensity is higher than in normal atomic MOTs, and a million times higher than\ncomparable red-detuned type-II MOTs, making it particularly attractive for\nmolecular MOTs which rely on type-II transitions. The loss of atoms from the\ntrap is dominated by ultracold collisions between Rb atoms. For typical\ntrapping conditions, we measure a loss rate of\n$1.8(4)\\times10^{-10}$cm$^{3}$s$^{-1}$.\n",
"title": "Blue-detuned magneto-optical trap"
}
| null | null | null | null | true | null |
6251
| null |
Default
| null | null |
null |
{
"abstract": " We prove that the Büchi topology and the automatic topology are Polish. We\nalso show that this cannot be fully extended to the case of a space of infinite\nlabelled binary trees; in particular the Büchi and the Muller topologies are\nnot Polish in this case.\n",
"title": "Polishness of some topologies related to word or tree automata"
}
| null | null |
[
"Computer Science",
"Mathematics"
] | null | true | null |
6252
| null |
Validated
| null | null |
null |
{
"abstract": " Obtaining accurate estimates of satellite drag coefficients in low Earth\norbit is a crucial component in positioning and collision avoidance. Simulators\ncan produce accurate estimates, but their computational expense is much too\nlarge for real-time application. A pilot study showed that Gaussian process\n(GP) surrogate models could accurately emulate simulations. However, cubic\nruntime for training GPs means that they could only be applied to a narrow\nrange of input configurations to achieve the desired level of accuracy. In this\npaper we show how extensions to the local approximate Gaussian Process (laGP)\nmethod allow accurate full-scale emulation. The new methodological\ncontributions, which involve a multi-level global/local modeling approach, and\na set-wise approach to local subset selection, are shown to perform well in\nbenchmark and synthetic data settings. We conclude by demonstrating that our\nmethod achieves the desired level of accuracy, besting simpler viable (i.e.,\ncomputationally tractable) global and local modeling approaches, when trained\non seventy thousand core hours of drag simulations for two real-world\nsatellites: the Hubble space telescope (HST) and the gravity recovery and\nclimate experiment (GRACE).\n",
"title": "Emulating satellite drag from large simulation experiments"
}
| null | null |
[
"Statistics"
] | null | true | null |
6253
| null |
Validated
| null | null |
null |
{
"abstract": " Cross-validation of predictive models is the de-facto standard for model\nselection and evaluation. In proper use, it provides an unbiased estimate of a\nmodel's predictive performance. However, data sets often undergo a preliminary\ndata-dependent transformation, such as feature rescaling or dimensionality\nreduction, prior to cross-validation. It is widely believed that such a\npreprocessing stage, if done in an unsupervised manner that does not consider\nthe class labels or response values, has no effect on the validity of\ncross-validation. In this paper, we show that this belief is not true.\nPreliminary preprocessing can introduce either a positive or negative bias into\nthe estimates of model performance. Thus, it may lead to sub-optimal choices of\nmodel parameters and invalid inference. In light of this, the scientific\ncommunity should re-examine the use of preliminary preprocessing prior to\ncross-validation across the various application domains. By default, all data\ntransformations, including unsupervised preprocessing stages, should be learned\nonly from the training samples, and then merely applied to the validation and\ntesting samples.\n",
"title": "Rescaling and other forms of unsupervised preprocessing introduce bias into cross-validation"
}
| null | null | null | null | true | null |
6254
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we deal with the problem of inferring causal directions when\nthe data is on discrete domain. By considering the distribution of the cause\n$P(X)$ and the conditional distribution mapping cause to effect $P(Y|X)$ as\nindependent random variables, we propose to infer the causal direction via\ncomparing the distance correlation between $P(X)$ and $P(Y|X)$ with the\ndistance correlation between $P(Y)$ and $P(X|Y)$. We infer \"$X$ causes $Y$\" if\nthe dependence coefficient between $P(X)$ and $P(Y|X)$ is smaller. Experiments\nare performed to show the performance of the proposed method.\n",
"title": "Causal Inference on Discrete Data via Estimating Distance Correlations"
}
| null | null | null | null | true | null |
6255
| null |
Default
| null | null |
null |
{
"abstract": " The discriminator of an integer sequence s = (s(i))_{i>=0}, introduced by\nArnold, Benkoski, and McCabe in 1985, is the function D_s(n) that sends n to\nthe least integer m such that the numbers s(0), s(1), ..., s(n-1) are pairwise\nincongruent modulo m. In this note we present a class of exponential sequences\nthat have the special property that their discriminators are shift-invariant,\ni.e., that the discriminator of the sequence is the same even if the sequence\nis shifted by any positive constant.\n",
"title": "A Class of Exponential Sequences with Shift-Invariant Discriminators"
}
| null | null | null | null | true | null |
6256
| null |
Default
| null | null |
null |
{
"abstract": " For a reductive group G defined over an algebraically closed field of\npositive characteristic, we show that the Frobenius contraction functor of\nG-modules is right adjoint to the Frobenius twist of the modules tensored with\nthe Steinberg module twice. It follows that the Frobenius contraction functor\npreserves injectivity, good filtrations, but not semisiplicity.\n",
"title": "Contraction par Frobenius et modules de Steinberg"
}
| null | null | null | null | true | null |
6257
| null |
Default
| null | null |
null |
{
"abstract": " We present an approach towards robust lane tracking for assisted and\nautonomous driving, particularly under poor visibility. Autonomous detection of\nlane markers improves road safety, and purely visual tracking is desirable for\nwidespread vehicle compatibility and reducing sensor intrusion, cost, and\nenergy consumption. However, visual approaches are often ineffective because of\na number of factors, including but not limited to occlusion, poor weather\nconditions, and paint wear-off. Our method, named SafeDrive, attempts to\nimprove visual lane detection approaches in drastically degraded visual\nconditions without relying on additional active sensors. In scenarios where\nvisual lane detection algorithms are unable to detect lane markers, the\nproposed approach uses location information of the vehicle to locate and access\nalternate imagery of the road and attempts detection on this secondary image.\nSubsequently, by using a combination of feature-based and pixel-based\nalignment, an estimated location of the lane marker is found in the current\nscene. We demonstrate the effectiveness of our system on actual driving data\nfrom locations in the United States with Google Street View as the source of\nalternate imagery.\n",
"title": "SafeDrive: A Robust Lane Tracking System for Autonomous and Assisted Driving Under Limited Visibility"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6258
| null |
Validated
| null | null |
null |
{
"abstract": " In this work we provide a couple of contributions to the analysis of\nlongitudinal data collected by smartphones in mobile health applications.\nFirst, we propose a novel statistical approach to disentangle personalized\ntreatment and \"time-of-the-day\" effects in observational studies. Under the\nassumption of no unmeasured confounders, we show how to use conditional\nindependence relations in the data in order to determine if a difference in\nperformance between activity tasks performed before and after the participant\nhas taken medication, are potentially due to an effect of the medication or to\na \"time-of-the-day\" effect (or still to both). Second, we show that smartphone\ndata collected from a given study participant can represent a \"digital\nfingerprint\" of the participant, and that classifiers of case/control labels,\nconstructed using longitudinal data, can show artificially improved performance\nwhen data from each participant is included in both training and test sets. We\nillustrate our contributions using data collected during the first 6 months of\nthe mPower study.\n",
"title": "On the analysis of personalized medication response and classification of case vs control patients in mobile health studies: the mPower case study"
}
| null | null |
[
"Statistics"
] | null | true | null |
6259
| null |
Validated
| null | null |
null |
{
"abstract": " The integrating factor and exponential time differencing methods are\nimplemented and tested for solving the time-dependent Kohn--Sham equations.\nPopular time propagation methods used in physics, as well as other robust\nnumerical approaches, are compared to these exponential integrator methods in\norder to judge the relative merit of the computational schemes. We determine an\nimprovement in accuracy of multiple orders of magnitude when describing\ndynamics driven primarily by a nonlinear potential. For cases of dynamics\ndriven by a time-dependent external potential, the accuracy of the exponential\nintegrator methods are less enhanced but still match or outperform the best of\nthe conventional methods tested.\n",
"title": "Exponential Integrators in Time-Dependent Density Functional Calculations"
}
| null | null | null | null | true | null |
6260
| null |
Default
| null | null |
null |
{
"abstract": " Control of multihop Wireless networks in a distributed manner while providing\nend-to-end delay requirements for different flows, is a challenging problem.\nUsing the notions of Draining Time and Discrete Review from the theory of fluid\nlimits of queues, an algorithm that meets delay requirements to various flows\nin a network is constructed. The algorithm involves an optimization which is\nimplemented in a cyclic distributed manner across nodes by using the technique\nof iterative gradient ascent, with minimal information exchange between nodes.\nThe algorithm uses time varying weights to give priority to flows. The\nperformance of the algorithm is studied in a network with interference modelled\nby independent sets.\n",
"title": "A Distributed Scheduling Algorithm to Provide Quality-of-Service in Multihop Wireless Networks"
}
| null | null | null | null | true | null |
6261
| null |
Default
| null | null |
null |
{
"abstract": " In a world of global trading, maritime safety, security and efficiency are\ncrucial issues. We propose a multi-task deep learning framework for vessel\nmonitoring using Automatic Identification System (AIS) data streams. We combine\nrecurrent neural networks with latent variable modeling and an embedding of AIS\nmessages to a new representation space to jointly address key issues to be\ndealt with when considering AIS data streams: massive amount of streaming data,\nnoisy data and irregular timesampling. We demonstrate the relevance of the\nproposed deep learning framework on real AIS datasets for a three-task setting,\nnamely trajectory reconstruction, anomaly detection and vessel type\nidentification.\n",
"title": "A Multi-task Deep Learning Architecture for Maritime Surveillance using AIS Data Streams"
}
| null | null | null | null | true | null |
6262
| null |
Default
| null | null |
null |
{
"abstract": " We describe the design of the CCI30 cryptocurrency index.\n",
"title": "The CCI30 Index"
}
| null | null | null | null | true | null |
6263
| null |
Default
| null | null |
null |
{
"abstract": " Third-generation neural networks, or Spiking Neural Networks (SNNs), aim at\nharnessing the energy efficiency of spike-domain processing by building on\ncomputing elements that operate on, and exchange, spikes. In this paper, the\nproblem of training a two-layer SNN is studied for the purpose of\nclassification, under a Generalized Linear Model (GLM) probabilistic neural\nmodel that was previously considered within the computational neuroscience\nliterature. Conventional classification rules for SNNs operate offline based on\nthe number of output spikes at each output neuron. In contrast, a novel\ntraining method is proposed here for a first-to-spike decoding rule, whereby\nthe SNN can perform an early classification decision once spike firing is\ndetected at an output neuron. Numerical results bring insights into the optimal\nparameter selection for the GLM neuron and on the accuracy-complexity trade-off\nperformance of conventional and first-to-spike decoding.\n",
"title": "Training Probabilistic Spiking Neural Networks with First-to-spike Decoding"
}
| null | null | null | null | true | null |
6264
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we propose the nonlinearity generation method to speed up and\nstabilize the training of deep convolutional neural networks. The proposed\nmethod modifies a family of activation functions as nonlinearity generators\n(NGs). NGs make the activation functions linear symmetric for their inputs to\nlower model capacity, and automatically introduce nonlinearity to enhance the\ncapacity of the model during training. The proposed method can be considered an\nunusual form of regularization: the model parameters are obtained by training a\nrelatively low-capacity model, that is relatively easy to optimize at the\nbeginning, with only a few iterations, and these parameters are reused for the\ninitialization of a higher-capacity model. We derive the upper and lower bounds\nof variance of the weight variation, and show that the initial symmetric\nstructure of NGs helps stabilize training. We evaluate the proposed method on\ndifferent frameworks of convolutional neural networks over two object\nrecognition benchmark tasks (CIFAR-10 and CIFAR-100). Experimental results\nshowed that the proposed method allows us to (1) speed up the convergence of\ntraining, (2) allow for less careful weight initialization, (3) improve or at\nleast maintain the performance of the model at negligible extra computational\ncost, and (4) easily train a very deep model.\n",
"title": "An Effective Training Method For Deep Convolutional Neural Network"
}
| null | null | null | null | true | null |
6265
| null |
Default
| null | null |
null |
{
"abstract": " Data-target association is an important step in multi-target localization for\nthe intelligent operation of un- manned systems in numerous applications such\nas search and rescue, traffic management and surveillance. The objective of\nthis paper is to present an innovative data association learning approach named\nmulti-layer K-means (MLKM) based on leveraging the advantages of some existing\nmachine learning approaches, including K-means, K-means++, and deep neural\nnetworks. To enable the accurate data association from different sensors for\nefficient target localization, MLKM relies on the clustering capabilities of\nK-means++ structured in a multi-layer framework with the error correction\nfeature that is motivated by the backpropogation that is well-known in deep\nlearning research. To show the effectiveness of the MLKM method, numerous\nsimulation examples are conducted to compare its performance with K-means,\nK-means++, and deep neural networks.\n",
"title": "A Multi-Layer K-means Approach for Multi-Sensor Data Pattern Recognition in Multi-Target Localization"
}
| null | null | null | null | true | null |
6266
| null |
Default
| null | null |
null |
{
"abstract": " In recent years, monaural speech separation has been formulated as a\nsupervised learning problem, which has been systematically researched and shown\nthe dramatical improvement of speech intelligibility and quality for human\nlisteners. However, it has not been well investigated whether the methods can\nbe employed as the front-end processing and directly improve the performance of\na machine listener, i.e., an automatic speech recognizer, without retraining or\njoint-training the acoustic model. In this paper, we explore the effectiveness\nof the independent front-end processing for the multi-conditional trained ASR\non the CHiME-3 challenge. We find that directly feeding the enhanced features\nto ASR can make 36.40% and 11.78% relative WER reduction for the GMM-based and\nDNN-based ASR respectively. We also investigate the affect of noisy phase and\ngeneralization ability under unmatched noise condition.\n",
"title": "Investigation of Monaural Front-End Processing for Robust ASR without Retraining or Joint-Training"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6267
| null |
Validated
| null | null |
null |
{
"abstract": " We investigate large-sample properties of treatment effect estimators under\nunknown interference in randomized experiments. The inferential target is a\ngeneralization of the average treatment effect estimand that marginalizes over\npotential spillover effects. We show that estimators commonly used to estimate\ntreatment effects under no-interference are consistent for the generalized\nestimand for several common experimental designs under limited but otherwise\narbitrary and unknown interference. The rates of convergence depend on the rate\nat which the amount of interference grows and the degree to which it aligns\nwith dependencies in treatment assignment. Importantly for practitioners, the\nresults imply that if one erroneously assumes that units do not interfere in a\nsetting with limited, or even moderate, interference, standard estimators are\nnevertheless likely to be close to an average treatment effect if the sample is\nsufficiently large.\n",
"title": "Average treatment effects in the presence of unknown interference"
}
| null | null | null | null | true | null |
6268
| null |
Default
| null | null |
null |
{
"abstract": " Nonlinear dynamical stochastic models are ubiquitous in different areas.\nExcitable media models are typical examples with large state dimensions. Their\nstatistical properties are often of great interest but are also very\nchallenging to compute. In this article, a theoretical framework to understand\nthe spatial localization for a large class of stochastically coupled nonlinear\nsystems in high dimensions is developed. Rigorous mathematical theories show\nthe covariance decay behavior due to both local and nonlocal effects, which\nresult from the diffusion and the mean field interaction, respectively. The\nanalysis is based on a comparison with an appropriate linear surrogate model,\nof which the covariance propagation can be computed explicitly. Two important\napplications of these theoretical results are discussed. They are the spatial\naveraging strategy for efficiently sampling the covariance matrix and the\nlocalization technique in data assimilation. Test examples of a surrogate\nlinear model and a stochastically coupled FitzHugh-Nagumo model for excitable\nmedia are adopted to validate the theoretical results. The latter is also used\nfor a systematical study of the spatial averaging strategy in efficiently\nsampling the covariance matrix in different dynamical regimes.\n",
"title": "Spatial localization for nonlinear dynamical stochastic models for excitable media"
}
| null | null | null | null | true | null |
6269
| null |
Default
| null | null |
null |
{
"abstract": " We provide a nonperturbative theory for photoionization of transparent\nsolids. By applying a particular steepest-descent method, we derive analytical\nexpressions for the photoionization rate within the two-band structure model,\nwhich consistently account for the $selection$ $rules$ related to the parity of\nthe number of absorbed photons ($odd$ or $even$). We demonstrate the crucial\nrole of the interference of the transition amplitudes (saddle-points), which in\nthe semi-classical limit, can be interpreted in terms of interfering quantum\ntrajectories. Keldysh's foundational work of laser physics [Sov. Phys. JETP 20,\n1307 (1965)] disregarded this interference, resulting in the violation of\n$selection$ $rules$. We provide an improved Keldysh photoionization theory and\nshow its excellent agreement with measurements for the frequency dependence of\nthe two-photon absorption and nonlinear refractive index coefficients in\ndielectrics.\n",
"title": "Nonlinear photoionization of transparent solids: a nonperturbative theory obeying selection rules"
}
| null | null | null | null | true | null |
6270
| null |
Default
| null | null |
null |
{
"abstract": " We focus on two particular aspects of model risk: the inability of a chosen\nmodel to fit observed market prices at a given point in time (calibration\nerror) and the model risk due to recalibration of model parameters (in\ncontradiction to the model assumptions). In this context, we follow the\napproach of Glasserman and Xu (2014) and use relative entropy as a pre-metric\nin order to quantify these two sources of model risk in a common framework, and\nconsider the trade-offs between them when choosing a model and the frequency\nwith which to recalibrate to the market. We illustrate this approach applied to\nthe models of Black and Scholes (1973) and Heston (1993), using option data for\nApple (AAPL) and Google (GOOG). We find that recalibrating a model more\nfrequently simply shifts model risk from one type to another, without any\nsubstantial reduction of aggregate model risk. Furthermore, moving to a more\ncomplicated stochastic model is seen to be counterproductive if one requires a\nhigh degree of robustness, for example as quantified by a 99 percent quantile\nof aggregate model risk.\n",
"title": "Quantifying the Model Risk Inherent in the Calibration and Recalibration of Option Pricing Models"
}
| null | null | null | null | true | null |
6271
| null |
Default
| null | null |
null |
{
"abstract": " Bright ring-like structure emission of the CN molecule has been observed in\nprotoplanetary disks. We investigate whether such structures are due to the\nmorphology of the disk itself or if they are instead an intrinsic feature of CN\nemission. With the intention of using CN as a diagnostic, we also address to\nwhich physical and chemical parameters CN is most sensitive. A set of disk\nmodels were run for different stellar spectra, masses, and physical structures\nvia the 2D thermochemical code DALI. An updated chemical network that accounts\nfor the most relevant CN reactions was adopted. Ring-shaped emission is found\nto be a common feature of all adopted models; the highest abundance is found in\nthe upper outer regions of the disk, and the column density peaks at 30-100 AU\nfor T Tauri stars with standard accretion rates. Higher mass disks generally\nshow brighter CN. Higher UV fields, such as those appropriate for T Tauri stars\nwith high accretion rates or for Herbig Ae stars or for higher disk flaring,\ngenerally result in brighter and larger rings. These trends are due to the main\nformation paths of CN, which all start with vibrationally excited H2*\nmolecules, that are produced through far ultraviolet (FUV) pumping of H2. The\nmodel results compare well with observed disk-integrated CN fluxes and the\nobserved location of the CN ring for the TW Hya disk. CN rings are produced\nnaturally in protoplanetary disks and do not require a specific underlying disk\nstructure such as a dust cavity or gap. The strong link between FUV flux and CN\nemission can provide critical information regarding the vertical structure of\nthe disk and the distribution of dust grains which affects the UV penetration,\nand could help to break some degeneracies in the SED fitting. In contrast with\nC2H or c-C3H2, the CN flux is not very sensitive to carbon and oxygen\ndepletion.\n",
"title": "CN rings in full protoplanetary disks around young stars as probes of disk structure"
}
| null | null | null | null | true | null |
6272
| null |
Default
| null | null |
null |
{
"abstract": " Measuring entity relatedness is a fundamental task for many natural language\nprocessing and information retrieval applications. Prior work often studies\nentity relatedness in static settings and an unsupervised manner. However,\nentities in real-world are often involved in many different relationships,\nconsequently entity-relations are very dynamic over time. In this work, we\npropose a neural networkbased approach for dynamic entity relatedness,\nleveraging the collective attention as supervision. Our model is capable of\nlearning rich and different entity representations in a joint framework.\nThrough extensive experiments on large-scale datasets, we demonstrate that our\nmethod achieves better results than competitive baselines.\n",
"title": "A Trio Neural Model for Dynamic Entity Relatedness Ranking"
}
| null | null | null | null | true | null |
6273
| null |
Default
| null | null |
null |
{
"abstract": " In this present study, we investigate solutions for fractional kinetic\nequations, involving k-Struve functions using Sumudu transform. The methodology\nand results can be considered and applied to various related fractional\nproblems in mathematical physics.\n",
"title": "Dynamic k-Struve Sumudu Solutions for Fractional Kinetic Equations"
}
| null | null | null | null | true | null |
6274
| null |
Default
| null | null |
null |
{
"abstract": " Following related work in law and policy, two notions of disparity have come\nto shape the study of fairness in algorithmic decision-making. Algorithms\nexhibit treatment disparity if they formally treat members of protected\nsubgroups differently; algorithms exhibit impact disparity when outcomes differ\nacross subgroups, even if the correlation arises unintentionally. Naturally, we\ncan achieve impact parity through purposeful treatment disparity. In one thread\nof technical work, papers aim to reconcile the two forms of parity proposing\ndisparate learning processes (DLPs). Here, the learning algorithm can see group\nmembership during training but produce a classifier that is group-blind at test\ntime. In this paper, we show theoretically that: (i) When other features\ncorrelate to group membership, DLPs will (indirectly) implement treatment\ndisparity, undermining the policy desiderata they are designed to address; (ii)\nWhen group membership is partly revealed by other features, DLPs induce\nwithin-class discrimination; and (iii) In general, DLPs provide a suboptimal\ntrade-off between accuracy and impact parity. Based on our technical analysis,\nwe argue that transparent treatment disparity is preferable to occluded methods\nfor achieving impact parity. Experimental results on several real-world\ndatasets highlight the practical consequences of applying DLPs vs. per-group\nthresholds.\n",
"title": "Does mitigating ML's impact disparity require treatment disparity?"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6275
| null |
Validated
| null | null |
null |
{
"abstract": " We study boundary conditions of topological sigma models with the goal of\ngeneralizing the concepts of anomalous symmetry and symmetry protected\ntopological order. We find a version of 't Hooft's anomaly matching conditions\non the renormalization group flow of boundaries of invertible topological sigma\nmodels and discuss several examples of anomalous boundary theories. We also\ncomment on bulk topological transitions in dynamical sigma models and argue\nthat one can, with care, use topological data to draw sigma model phase\ndiagrams.\n",
"title": "Topological Terms and Phases of Sigma Models"
}
| null | null | null | null | true | null |
6276
| null |
Default
| null | null |
null |
{
"abstract": " We prove the pointwise decay of solutions to three linear equations: (i) the\ntransport equation in phase space generalizing the classical Vlasov equation,\n(ii) the linear Schrodinger equation, (iii) the Airy (linear KdV) equation. The\nusual proofs use explicit representation formulae, and either obtain\n$L^1$---$L^\\infty$ decay through directly estimating the fundamental solution\nin physical space, or by studying oscillatory integrals coming from the\nrepresentation in Fourier space. Our proof instead combines \"vector field\"\ncommutators that capture the inherent symmetries of the relevant equations with\nconservation laws for mass and energy to get space-time weighted energy\nestimates. Combined with a simple version of Sobolev's inequality this gives\npointwise decay as desired. In the case of the Vlasov and Schrodinger equations\nwe can recover sharp pointwise decay; in the Schrodinger case we also show how\nto obtain local energy decay as well as Strichartz-type estimates. For the Airy\nequation we obtain a local energy decay that is almost sharp from the scaling\npoint of view, but nonetheless misses the classical estimates by a gap. This\nwork is inspired by the work of Klainerman on $L^2$---$L^\\infty$ decay of wave\nequations, as well as the recent work of Fajman, Joudioux, and Smulevici on\ndecay of mass distributions for the relativistic Vlasov equation.\n",
"title": "A commuting-vector-field approach to some dispersive estimates"
}
| null | null | null | null | true | null |
6277
| null |
Default
| null | null |
null |
{
"abstract": " Answering a problem posed by the second author on Mathoverflow, we prove that\nthe connected countable Hausdorff spaces constructed by Bing and Ritter are\ntopologically homogeneous.\n",
"title": "The connected countable spaces of Bing and Ritter are topologically homogeneous"
}
| null | null | null | null | true | null |
6278
| null |
Default
| null | null |
null |
{
"abstract": " Many different classification tasks need to manage structured data, which are\nusually modeled as graphs. Moreover, these graphs can be dynamic, meaning that\nthe vertices/edges of each graph may change during time. Our goal is to jointly\nexploit structured data and temporal information through the use of a neural\nnetwork model. To the best of our knowledge, this task has not been addressed\nusing these kind of architectures. For this reason, we propose two novel\napproaches, which combine Long Short-Term Memory networks and Graph\nConvolutional Networks to learn long short-term dependencies together with\ngraph structure. The quality of our methods is confirmed by the promising\nresults achieved.\n",
"title": "Dynamic Graph Convolutional Networks"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6279
| null |
Validated
| null | null |
null |
{
"abstract": " An atomic force microscope (AFM) is capable of producing ultra-high\nresolution measurements of nanoscopic objects and forces. It is an\nindispensable tool for various scientific disciplines such as molecular\nengineering, solid-state physics, and cell biology. Prior to a given\nexperiment, the AFM must be calibrated by fitting a spectral density model to\nbaseline recordings. However, since AFM experiments typically collect large\namounts of data, parameter estimation by maximum likelihood can be\nprohibitively expensive. Thus, practitioners routinely employ a much faster\nleast-squares estimation method, at the cost of substantially reduced\nstatistical efficiency. Additionally, AFM data is often contaminated by\nperiodic electronic noise, to which parameter estimates are highly sensitive.\nThis article proposes a two-stage estimator to address these issues.\nPreliminary parameter estimates are first obtained by a variance-stabilizing\nprocedure, by which the simplicity of least-squares combines with the\nefficiency of maximum likelihood. A test for spectral periodicities then\neliminates high-impact outliers, considerably and robustly protecting the\nsecond-stage estimator from the effects of electronic noise. Simulation and\nexperimental results indicate that a two- to ten-fold reduction in mean squared\nerror can be expected by applying our methodology.\n",
"title": "Robust and Efficient Parametric Spectral Estimation in Atomic Force Microscopy"
}
| null | null | null | null | true | null |
6280
| null |
Default
| null | null |
null |
{
"abstract": " Considering its advantages in dealing with high-dimensional visual input and\nlearning control policies in discrete domain, Deep Q Network (DQN) could be an\nalternative method of traditional auto-focus means in the future. In this\npaper, based on Deep Reinforcement Learning, we propose an end-to-end approach\nthat can learn auto-focus policies from visual input and finish at a clear spot\nautomatically. We demonstrate that our method - discretizing the action space\nwith coarse to fine steps and applying DQN is not only a solution to auto-focus\nbut also a general approach towards vision-based control problems. Separate\nphases of training in virtual and real environments are applied to obtain an\neffective model. Virtual experiments, which are carried out after the virtual\ntraining phase, indicates that our method could achieve 100% accuracy on a\ncertain view with different focus range. Further training on real robots could\neliminate the deviation between the simulator and real scenario, leading to\nreliable performances in real applications.\n",
"title": "A Robotic Auto-Focus System based on Deep Reinforcement Learning"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6281
| null |
Validated
| null | null |
null |
{
"abstract": " We extend the approach of wall modeling via function enrichment to\ndetached-eddy simulation. The wall model aims at using coarse cells in the\nnear-wall region by modeling the velocity profile in the viscous sublayer and\nlog-layer. However, unlike other wall models, the full Navier-Stokes equations\nare still discretely fulfilled, including the pressure gradient and convective\nterm. This is achieved by enriching the elements of the high-order\ndiscontinuous Galerkin method with the law-of-the-wall. As a result, the\nGalerkin method can \"choose\" the optimal solution among the polynomial and\nenrichment shape functions. The detached-eddy simulation methodology provides a\nsuitable turbulence model for the coarse near-wall cells. The approach is\napplied to wall-modeled LES of turbulent channel flow in a wide range of\nReynolds numbers. Flow over periodic hills shows the superiority compared to an\nequilibrium wall model under separated flow conditions.\n",
"title": "Wall modeling via function enrichment: extension to detached-eddy simulation"
}
| null | null | null | null | true | null |
6282
| null |
Default
| null | null |
null |
{
"abstract": " The aim of this work is to study the existence of a periodic solutions of\nnth-order differential equations with delay d dt x(t) + d 2 dt 2 x(t) + d 3 dt\n3 x(t) + ... + d n dt n x(t) = Ax(t) + L(xt) + f (t). Our approach is based on\nthe M-boundedness of linear operators, Fourier type, B s p,q-multipliers and\nBesov spaces.\n",
"title": "Existence and uniqueness of periodic solution of nth-order Equations with delay in Banach space having Fourier type"
}
| null | null | null | null | true | null |
6283
| null |
Default
| null | null |
null |
{
"abstract": " This paper proposes a totally constructive approach for the proof of\nHilbert's theorem on ternary quartic forms. The main contribution is the ladder\ntechnique, with which the Hilbert's theorem is proved vividly.\n",
"title": "A proof of Hilbert's theorem on ternary quartic forms with the ladder technique"
}
| null | null | null | null | true | null |
6284
| null |
Default
| null | null |
null |
{
"abstract": " According to a report online, more than 200 million unique users search for\njobs online every month. This incredibly large and fast growing demand has\nenticed software giants such as Google and Facebook to enter this space, which\nwas previously dominated by companies such as LinkedIn, Indeed and\nCareerBuilder. Recently, Google released their \"AI-powered Jobs Search Engine\",\n\"Google For Jobs\" while Facebook released \"Facebook Jobs\" within their\nplatform. These current job search engines and platforms allow users to search\nfor jobs based on general narrow filters such as job title, date posted,\nexperience level, company and salary. However, they have severely limited\nfilters relating to skill sets such as C++, Python, and Java and company\nrelated attributes such as employee size, revenue, technographics and\nmicro-industries. These specialized filters can help applicants and companies\nconnect at a very personalized, relevant and deeper level. In this paper we\npresent a framework that provides an end-to-end \"Data-driven Jobs Search\nEngine\". In addition, users can also receive potential contacts of recruiters\nand senior positions for connection and networking opportunities. The high\nlevel implementation of the framework is described as follows: 1) Collect job\npostings data in the United States, 2) Extract meaningful tokens from the\npostings data using ETL pipelines, 3) Normalize the data set to link company\nnames to their specific company websites, 4) Extract and ranking the skill\nsets, 5) Link the company names and websites to their respective company level\nattributes with the EVERSTRING Company API, 6) Run user-specific search queries\non the database to identify relevant job postings and 7) Rank the job search\nresults. This framework offers a highly customizable and highly targeted search\nexperience for end users.\n",
"title": "Data-driven Job Search Engine Using Skills and Company Attribute Filters"
}
| null | null | null | null | true | null |
6285
| null |
Default
| null | null |
null |
{
"abstract": " A Leonard pair is a pair of diagonalizable linear transformations of a\nfinite-dimensional vector space, each of which acts in an irreducible\ntridiagonal fashion on an eigenbasis for the other one. Let $\\mathbb F$ denote\nan algebraically closed field, and fix a nonzero $q \\in \\mathbb F$ that is not\na root of unity. The universal double affine Hecke algebra (DAHA) $\\hat{H}_q$\nof type $(C_1^\\vee,C_1)$ is the associative $\\mathbb F$-algebra defined by\ngenerators $\\lbrace t_i^{\\pm 1}\\rbrace_{i=0}^3$ and relations (i)\n$t_it_i^{-1}=t_i^{-1}t_i=1$; (ii) $t_i+t_i^{-1}$ is central; (iii)\n$t_0t_1t_2t_3 = q^{-1}$. We consider the elements $X=t_3t_0$ and $Y=t_0t_1$ of\n$\\hat{H}_q$. Let $\\mathcal V$ denote a finite-dimensional irreducible\n$\\hat{H}_q$-module on which each of $X$, $Y$ is diagonalizable and $t_0$ has\ntwo distinct eigenvalues. Then $\\mathcal V$ is a direct sum of the two\neigenspaces of $t_0$. We show that the pair $X+X^{-1}$, $Y+Y^{-1}$ acts on each\neigenspace as a Leonard pair, and each of these Leonard pairs falls into a\nclass said to have $q$-Racah type. Thus from $\\mathcal V$ we obtain a pair of\nLeonard pairs of $q$-Racah type. It is known that a Leonard pair of $q$-Racah\ntype is determined up to isomorphism by a parameter sequence $(a,b,c,d)$ called\nits Huang data. Given a pair of Leonard pairs of $q$-Racah type, we find\nnecessary and sufficient conditions on their Huang data for that pair to come\nfrom the above construction.\n",
"title": "The universal DAHA of type $(C_1^\\vee,C_1)$ and Leonard pairs of $q$-Racah type"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6286
| null |
Validated
| null | null |
null |
{
"abstract": " The rise and fall of artificial neural networks is well documented in the\nscientific literature of both computer science and computational chemistry. Yet\nalmost two decades later, we are now seeing a resurgence of interest in deep\nlearning, a machine learning algorithm based on multilayer neural networks.\nWithin the last few years, we have seen the transformative impact of deep\nlearning in many domains, particularly in speech recognition and computer\nvision, to the extent that the majority of expert practitioners in those field\nare now regularly eschewing prior established models in favor of deep learning\nmodels. In this review, we provide an introductory overview into the theory of\ndeep neural networks and their unique properties that distinguish them from\ntraditional machine learning algorithms used in cheminformatics. By providing\nan overview of the variety of emerging applications of deep neural networks, we\nhighlight its ubiquity and broad applicability to a wide range of challenges in\nthe field, including QSAR, virtual screening, protein structure prediction,\nquantum chemistry, materials design and property prediction. In reviewing the\nperformance of deep neural networks, we observed a consistent outperformance\nagainst non-neural networks state-of-the-art models across disparate research\ntopics, and deep neural network based models often exceeded the \"glass ceiling\"\nexpectations of their respective tasks. Coupled with the maturity of\nGPU-accelerated computing for training deep neural networks and the exponential\ngrowth of chemical data on which to train these networks on, we anticipate that\ndeep learning algorithms will be a valuable tool for computational chemistry.\n",
"title": "Deep Learning for Computational Chemistry"
}
| null | null | null | null | true | null |
6287
| null |
Default
| null | null |
null |
{
"abstract": " Most approaches in algorithmic fairness constrain machine learning methods so\nthe resulting predictions satisfy one of several intuitive notions of fairness.\nWhile this may help private companies comply with non-discrimination laws or\navoid negative publicity, we believe it is often too little, too late. By the\ntime the training data is collected, individuals in disadvantaged groups have\nalready suffered from discrimination and lost opportunities due to factors out\nof their control. In the present work we focus instead on interventions such as\na new public policy, and in particular, how to maximize their positive effects\nwhile improving the fairness of the overall system. We use causal methods to\nmodel the effects of interventions, allowing for potential interference--each\nindividual's outcome may depend on who else receives the intervention. We\ndemonstrate this with an example of allocating a budget of teaching resources\nusing a dataset of schools in New York City.\n",
"title": "Causal Interventions for Fairness"
}
| null | null | null | null | true | null |
6288
| null |
Default
| null | null |
null |
{
"abstract": " Predictive geometric models deliver excellent results for many Machine\nLearning use cases. Despite their undoubted performance, neural predictive\nalgorithms can show unexpected degrees of instability and variance,\nparticularly when applied to large datasets. We present an approach to measure\nchanges in geometric models with respect to both output consistency and\ntopological stability. Considering the example of a recommender system using\nword2vec, we analyze the influence of single data points, approximation methods\nand parameter settings. Our findings can help to stabilize models where needed\nand to detect differences in informational value of data points on a large\nscale.\n",
"title": "Analyzing Hypersensitive AI: Instability in Corporate-Scale Machine Learning"
}
| null | null | null | null | true | null |
6289
| null |
Default
| null | null |
null |
{
"abstract": " Almost a decade has passed since the serendipitous discovery of the\niron-based high temperature superconductors (FeSCs) in 2008. The question of\nhow much similarity the FeSCs have with the copper oxide high temperature\nsuperconductors emerged since the initial discovery of long-range\nantiferromagnetism in the FeSCs in proximity to superconductivity. Despite the\ngreat resemblance in their phase diagrams, there exist important disparities\nbetween FeSCs and cuprates that need to be considered in order to paint a full\npicture of these two families of high temperature superconductors. One of the\nkey differences lies in the multi-orbital multi-band nature of FeSCs, in\ncontrast to the effective single-band model for cuprates. Due to the complexity\nof multi-orbital band structures, the orbital degree of freedom is often\nneglected in formulating the theoretical models for FeSCs. On the experimental\nside, systematic studies of the orbital related phenomena in FeSCs have been\nlargely lacking. In this review, we summarize angle-resolved photoemission\nspectroscopy (ARPES) measurements across various FeSC families in literature,\nfocusing on the systematic trend of orbital dependent electron correlations and\nthe role of different Fe 3d orbitals in driving the nematic transition, the\nspin-density-wave transition, and implications for superconductivity.\n",
"title": "Role of the orbital degree of freedom in iron-based superconductors"
}
| null | null |
[
"Physics"
] | null | true | null |
6290
| null |
Validated
| null | null |
null |
{
"abstract": " Perpetual points (PPs) are special critical points for which the magnitude of\nacceleration describing dynamics drops to zero, while the motion is still\npossible (stationary points are excluded), e.g. considering the motion of the\nparticle in the potential field, at perpetual point it has zero acceleration\nand non-zero velocity. We show that using PPs we can trace all the stable fixed\npoints in the system, and that the structure of trajectories leading from\nformer points to stable equilibria may be similar to orbits obtained from\nunstable stationary points. Moreover, we argue that the concept of perpetual\npoints may be useful in tracing unexpected attractors (hidden or rare\nattractors with small basins of attraction). We show potential applicability of\nthis approach by analysing several representative systems of physical\nsignificance, including the damped oscillator, pendula and the Henon map. We\nsuggest that perpetual points may be a useful tool for localization of\nco-existing attractors in dynamical systems.\n",
"title": "Perpetual points: New tool for localization of co-existing attractors in dynamical systems"
}
| null | null | null | null | true | null |
6291
| null |
Default
| null | null |
null |
{
"abstract": " A squared error loss remains the most commonly used loss function for\nconstructing a Bayes estimator of the parameter of interest. It, however, can\nlead to sub-optimal solutions when a parameter is defined on a restricted\nspace. It can also be an inappropriate choice in the context when an\noverestimation and/or underestimation results in severe consequences and a more\nconservative estimator is preferred. We advocate a class of loss functions for\nparameters defined on restricted spaces which infinitely penalize boundary\ndecisions like the squared error loss does on the real line. We also recall\nseveral properties of loss functions such as symmetry, convexity and\ninvariance. We propose generalizations of the squared error loss function for\nparameters defined on the positive real line and on an interval. We provide\nexplicit solutions for corresponding Bayes estimators and discuss multivariate\nextensions. Three well-known Bayesian estimation problems are used to\ndemonstrate inferential benefits the novel Bayes estimators can provide in the\ncontext of restricted estimation.\n",
"title": "Loss Functions in Restricted Parameter Spaces and Their Bayesian Applications"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
6292
| null |
Validated
| null | null |
null |
{
"abstract": " The study of complex systems benefits from graph models and their analysis.\nIn particular, the eigendecomposition of the graph Laplacian lets emerge\nproperties of global organization from local interactions; e.g., the Fiedler\nvector has the smallest non-zero eigenvalue and plays a key role for graph\nclustering. Graph signal processing focusses on the analysis of signals that\nare attributed to the graph nodes. The eigendecomposition of the graph\nLaplacian allows to define the graph Fourier transform and extend conventional\nsignal-processing operations to graphs. Here, we introduce the design of\nSlepian graph signals, by maximizing energy concentration in a predefined\nsubgraph for a graph spectral bandlimit. We establish a novel link with\nclassical Laplacian embedding and graph clustering, which provides a meaning to\nlocalized graph frequencies.\n",
"title": "When Slepian Meets Fiedler: Putting a Focus on the Graph Spectrum"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6293
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper, a scheme for the encryption and decryption of colored images\nby using the Lorenz system and the discrete cosine transform in two dimensions\n(DCT2) is proposed. Although chaos is random, it has deterministic features\nthat can be used for encryption; further, the same sequences can be produced at\nthe transmitter and receiver under the same initial conditions. Another\nproperty of DCT2 is that the energy is concentrated in some elements of the\ncoefficients. These two properties are used to efficiently encrypt and recover\nthe image at the receiver by using three different keys with three different\npredefined number of shifts for each instance of key usage. Simulation results\nand statistical analysis show that the scheme high performance in weakening the\ncorrelation between the pixels of the image that resulted from the inverse of\nhighest energy values of DCT2 that form 99.9 % of the energy as well as those\nof the difference image.\n",
"title": "Colored Image Encryption and Decryption Using Chaotic Lorenz System and DCT2"
}
| null | null | null | null | true | null |
6294
| null |
Default
| null | null |
null |
{
"abstract": " Given a $4$-manifold $\\hat{M}$ and two homeomorphic surfaces $\\Sigma_1,\n\\Sigma_2$ smoothly embedded in $\\hat{M}$ with genus more than 1, we remove the\nneighborhoods of the surfaces and obtain a new $4$-manifold $M$ from gluing two\nboundaries $S^1 \\times \\Sigma_1$ and $S^1 \\times \\Sigma_1.$ In this artice, we\nprove a gluing formula which describes the relation of the Seiberg-Witten\ninvariants of $M$ and $\\hat{M}.$ Moreover, we show the application of the\nformula on the existence condition of the symplectic structure on a family of\n$4$-manfolds under some conditions.\n",
"title": "Self-Gluing formula of the monopole invariant and its application"
}
| null | null | null | null | true | null |
6295
| null |
Default
| null | null |
null |
{
"abstract": " This report describes the development of an aptamer for sensing azole\nantifungal drugs for therapeutic drug monitoring. Modified Synthetic Evolution\nof Ligands through Exponential Enrichment (SELEX) was used to discover a DNA\naptamer recognizing azole class antifungal drugs. This aptamer undergoes a\nsecondary structural change upon binding to its target molecule as shown\nthrough fluorescence anisotropy-based binding measurements. Experiments using\ncircular dichroism spectroscopy, revealed a unique double G-quadruplex\nstructure that was essential and specific for binding to the azole antifungal\ntarget. Aptamer-functionalized Graphene Field Effect Transistor (GFET) devices\nwere created and used to measure the binding of strength of azole antifungals\nto this surface. In total this aptamer and the supporting sensing platform\ncould provide a valuable tool for improving the treatment of patients with\ninvasive fungal infections.\n",
"title": "An aptamer-biosensor for azole class antifungal drugs"
}
| null | null | null | null | true | null |
6296
| null |
Default
| null | null |
null |
{
"abstract": " In the paper \"Einstein metrics on compact simple Lie groups attached to\nstandard triples\", the authors introduced the definition of standard triples\nand proved that every compact simple Lie group $G$ attached to a standard\ntriple $(G,K,H)$ admits a left-invariant Einstein metric which is not naturally\nreductive except the standard triple $(\\Sp(4),2\\Sp(2),4\\Sp(1))$. For the triple\n$(\\Sp(4),2\\Sp(2),4\\Sp(1))$, we find there exists an involution pair of $\\sp(4)$\nsuch that $4\\sp(1)$ is the fixed point of the pair, and then give the\ndecomposition of $\\sp(4)$ as a direct sum of irreducible\n$\\ad(4\\sp(1))$-modules. But $\\Sp(4)/4\\Sp(1)$ is not a generalized Wallach\nspace. Furthermore we give left-invariant Einstein metrics on $\\Sp(4)$ which\nare non-naturally reductive and $\\Ad(4\\Sp(1))$-invariant. For the general case\n$(\\Sp(2n_1n_2),2\\Sp(n_1n_2),2n_2\\Sp(n_1))$, there exist $2n_2-1$ involutions of\n$\\sp(2n_1n_2)$ such that $2n_2\\sp(n_1))$ is the fixed point of these $2n_2-1$\ninvolutions, and it follows the decomposition of $\\sp(2n_1n_2)$ as a direct sum\nof irreducible $\\ad(2n_2\\sp(n_1))$-modules. In order to give new non-naturally\nreductive and $\\Ad(2n_2\\Sp(n_1)))$-invariant Einstein metrics on\n$\\Sp(2n_1n_2)$, we prove a general result, i.e. $\\Sp(2k+l)$ admits at least two\nnon-naturally reductive Einstein metrics which are\n$\\Ad(\\Sp(k)\\times\\Sp(k)\\times\\Sp(l))$-invariant if $k<l$. It implies that every\ncompact simple Lie group $\\Sp(n)$ for $n\\geq 4$ admits at least\n$2[\\frac{n-1}{3}]$ non-naturally reductive left-invariant Einstein metrics.\n",
"title": "Notes on \"Einstein metrics on compact simple Lie groups attached to standard triples\""
}
| null | null | null | null | true | null |
6297
| null |
Default
| null | null |
null |
{
"abstract": " We prove that a quasi-isometric map, and more generally a coarse embedding,\nbetween pinched Hadamard manifolds is within bounded distance from a unique\nharmonic map.\n",
"title": "Harmonic quasi-isometric maps II : negatively curved manifolds"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6298
| null |
Validated
| null | null |
null |
{
"abstract": " The supercomputing platforms available for high performance computing based\nresearch evolve at a great rate. However, this rapid development of novel\ntechnologies requires constant adaptations and optimizations of the existing\ncodes for each new machine architecture. In such context, minimizing time of\nefficiently porting the code on a new platform is of crucial importance. A\npossible solution for this common challenge is to use simulations of the\napplication that can assist in detecting performance bottlenecks. Due to\nprohibitive costs of classical cycle-accurate simulators, coarse-grain\nsimulations are more suitable for large parallel and distributed systems. We\npresent a procedure of implementing the profiling for openQCD code [1] through\nsimulation, which will enable the global reduction of the cost of profiling and\noptimizing this code commonly used in the lattice QCD community. Our approach\nis based on well-known SimGrid simulator [2], which allows for fast and\naccurate performance predictions of HPC codes. Additionally, accurate\nestimations of the program behavior on some future machines, not yet accessible\nto us, are anticipated.\n",
"title": "Platform independent profiling of a QCD code"
}
| null | null | null | null | true | null |
6299
| null |
Default
| null | null |
null |
{
"abstract": " Deep learning (DL) defines a new data-driven programming paradigm that\nconstructs the internal system logic of a crafted neuron network through a set\nof training data. We have seen wide adoption of DL in many safety-critical\nscenarios. However, a plethora of studies have shown that the state-of-the-art\nDL systems suffer from various vulnerabilities which can lead to severe\nconsequences when applied to real-world applications. Currently, the testing\nadequacy of a DL system is usually measured by the accuracy of test data.\nConsidering the limitation of accessible high quality test data, good accuracy\nperformance on test data can hardly provide confidence to the testing adequacy\nand generality of DL systems. Unlike traditional software systems that have\nclear and controllable logic and functionality, the lack of interpretability in\na DL system makes system analysis and defect detection difficult, which could\npotentially hinder its real-world deployment. In this paper, we propose\nDeepGauge, a set of multi-granularity testing criteria for DL systems, which\naims at rendering a multi-faceted portrayal of the testbed. The in-depth\nevaluation of our proposed testing criteria is demonstrated on two well-known\ndatasets, five DL systems, and with four state-of-the-art adversarial attack\ntechniques against DL. The potential usefulness of DeepGauge sheds light on the\nconstruction of more generic and robust DL systems.\n",
"title": "DeepGauge: Multi-Granularity Testing Criteria for Deep Learning Systems"
}
| null | null | null | null | true | null |
6300
| null |
Default
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.