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": " We consider maximum likelihood estimation for Gaussian Mixture Models (Gmms).\nThis task is almost invariably solved (in theory and practice) via the\nExpectation Maximization (EM) algorithm. EM owes its success to various\nfactors, of which is its ability to fulfill positive definiteness constraints\nin closed form is of key importance. We propose an alternative to EM by\nappealing to the rich Riemannian geometry of positive definite matrices, using\nwhich we cast Gmm parameter estimation as a Riemannian optimization problem.\nSurprisingly, such an out-of-the-box Riemannian formulation completely fails\nand proves much inferior to EM. This motivates us to take a closer look at the\nproblem geometry, and derive a better formulation that is much more amenable to\nRiemannian optimization. We then develop (Riemannian) batch and stochastic\ngradient algorithms that outperform EM, often substantially. We provide a\nnon-asymptotic convergence analysis for our stochastic method, which is also\nthe first (to our knowledge) such global analysis for Riemannian stochastic\ngradient. Numerous empirical results are included to demonstrate the\neffectiveness of our methods.\n",
"title": "An Alternative to EM for Gaussian Mixture Models: Batch and Stochastic Riemannian Optimization"
}
| null | null | null | null | true | null |
8901
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we introduce and study the coprime quantum chain, i.e. a\nstrongly correlated quantum system defined in terms of the integer eigenvalues\n$n_i$ of the occupation number operators at each site of a chain of length $M$.\nThe $n_i$'s take value in the interval $[2,q]$ and may be regarded as $S_z$\neigenvalues in the spin representation $j = (q-2)/2$. The distinctive\ninteraction of the model is based on the coprimality matrix $\\bf \\Phi$: for the\nferromagnetic case, this matrix assigns lower energy to configurations where\noccupation numbers $n_i$ and $n_{i+1}$ of neighbouring sites share a common\ndivisor, while for the anti-ferromagnetic case it assigns lower energy to\nconfigurations where $n_i$ and $n_{i+1}$ are coprime. The coprime chain, both\nin the ferro and anti-ferromagnetic cases, may present an exponential number of\nground states whose values can be exactly computed by means of graph\ntheoretical tools. In the ferromagnetic case there are generally also\nfrustration phenomena. A fine tuning of local operators may lift the\nexponential ground state degeneracy and, according to which operators are\nswitched on, the system may be driven into different classes of universality,\namong which the Ising or Potts universality class. The paper also contains an\nappendix by Don Zagier on the exact eigenvalues and eigenvectors of the\ncoprimality matrix in the limit $q \\rightarrow \\infty$.\n",
"title": "The Coprime Quantum Chain"
}
| null | null | null | null | true | null |
8902
| null |
Default
| null | null |
null |
{
"abstract": " Kraichnan seminal ideas on inverse cascades yielded new tools to study common\nphenomena in geophysical turbulent flows. In the atmosphere and the oceans,\nrotation and stratification result in a flow that can be approximated as\ntwo-dimensional at very large scales, but which requires considering\nthree-dimensional effects to fully describe turbulent transport processes and\nnon-linear phenomena. Motions can thus be classified into two classes: fast\nmodes consisting of inertia-gravity waves, and slow quasi-geostrophic modes for\nwhich the Coriolis force and horizontal pressure gradients are close to\nbalance. In this paper we review previous results on the strength of the\ninverse cascade in rotating and stratified flows, and then present new results\non the effect of varying the strength of rotation and stratification (measured\nby the ratio $N/f$ of the Brunt-Väisäla frequency to the Coriolis\nfrequency) on the amplitude of the waves and on the flow quasi-geostrophic\nbehavior. We show that the inverse cascade is more efficient in the range of\n$N/f$ for which resonant triads do not exist, $1/2 \\le N/f \\le 2$. We then use\nthe spatio-temporal spectrum, and characterization of the flow temporal and\nspatial scales, to show that in this range slow modes dominate the dynamics,\nwhile the strength of the waves (and their relevance in the flow dynamics) is\nweaker.\n",
"title": "Inverse cascades and resonant triads in rotating and stratified turbulence"
}
| null | null | null | null | true | null |
8903
| null |
Default
| null | null |
null |
{
"abstract": " A proposal to improve routing security---Route Origin Authorization\n(ROA)---has been standardized. A ROA specifies which network is allowed to\nannounce a set of Internet destinations. While some networks now specify ROAs,\nlittle is known about whether other networks check routes they receive against\nthese ROAs, a process known as Route Origin Validation (ROV). Which networks\nblindly accept invalid routes? Which reject them outright? Which de-preference\nthem if alternatives exist?\nRecent analysis attempts to use uncontrolled experiments to characterize ROV\nadoption by comparing valid routes and invalid routes. However, we argue that\ngaining a solid understanding of ROV adoption is impossible using currently\navailable data sets and techniques. Our measurements suggest that, although\nsome ISPs are not observed using invalid routes in uncontrolled experiments,\nthey are actually using different routes for (non-security) traffic engineering\npurposes, without performing ROV. We conclude with a description of a\ncontrolled, verifiable methodology for measuring ROV and present three ASes\nthat do implement ROV, confirmed by operators.\n",
"title": "Towards a Rigorous Methodology for Measuring Adoption of RPKI Route Validation and Filtering"
}
| null | null | null | null | true | null |
8904
| null |
Default
| null | null |
null |
{
"abstract": " We present a primal--dual memory efficient algorithm for solving a relaxed\nversion of the general transportation problem. Our approach approximates the\noriginal cost function with a differentiable one that is solved as a sequence\nof weighted quadratic transportation problems. The new formulation allows us to\nsolve differentiable, non-- convex transportation problems.\n",
"title": "Half-quadratic transportation problems"
}
| null | null | null | null | true | null |
8905
| null |
Default
| null | null |
null |
{
"abstract": " We establish a conceptual framework for the identification and the\ncharacterization of induced interactions in binary mixtures and reveal their\nintricate relation to entanglement between the components or species of the\nmixture. Exploiting an expansion in terms of the strength of the entanglement\namong the two species, enables us to deduce an effective single-species\ndescription. In this way, we naturally incorporate the mutual feedback of the\nspecies and obtain induced interactions for both species which are effectively\npresent among the particles of same type. Importantly, our approach\nincorporates few-body and inhomogeneous systems extending the scope of induced\ninteractions where two particles interact via a bosonic bath-type environment.\nEmploying the example of a one-dimensional spin-polarized ultracold Bose-Fermi\nmixture, we obtain induced Bose-Bose and Fermi-Fermi interactions with\nshort-range attraction and long-range repulsion. With this, we show how beyond\nspecies mean-field physics visible in the two-body correlation functions can be\nunderstood via the induced interactions.\n",
"title": "Entanglement induced interactions in binary mixtures"
}
| null | null | null | null | true | null |
8906
| null |
Default
| null | null |
null |
{
"abstract": " The fiducial is not unique in general, but we prove that in a restricted\nclass of models it is uniquely determined by the sampling distribution of the\ndata. It depends in particular not on the choice of a data generating model.\nThe arguments lead to a generalization of the classical formula found by Fisher\n(1930). The restricted class includes cases with discrete distributions, the\ncase of the shape parameter in the Gamma distribution, and also the case of the\ncorrelation coefficient in a bivariate Gaussian model. One of the examples can\nalso be used in a pedagogical context to demonstrate possible difficulties with\nlikelihood-, Bayesian-, and bootstrap-inference. Examples that demonstrate\nnon-uniqueness are also presented. It is explained that they can be seen as\ncases with restrictions on the parameter space. Motivated by this the concept\nof a conditional fiducial model is introduced. This class of models includes\nthe common case of iid samples from a one-parameter model investigated by\nHannig (2013), the structural group models investigated by Fraser (1968), and\nalso certain models discussed by Fisher (1973) in his final writing on the\nsubject.\n",
"title": "Conditional fiducial models"
}
| null | null | null | null | true | null |
8907
| null |
Default
| null | null |
null |
{
"abstract": " Statistical thinking partially depends upon an iterative process by which\nessential features of a problem setting are identified and mapped onto an\nabstract model or archetype, and then translated back into the context of the\noriginal problem setting (Wild and Pfannkuch 1999). Assessment in introductory\nstatistics often relies on tasks that present students with data in context and\nexpects them to choose and describe an appropriate model. This study explores\npost-secondary student responses to an alternative task that prompts students\nto clearly identify a sample, population, statistic, and parameter using a\ncontext of their own invention. The data include free text narrative responses\nof a random sample of 500 students from a sample of more than 1600 introductory\nstatistics students. Results suggest that students' responses often portrayed\nsample and population accurately. Portrayals of statistic and parameter were\nless reliable and were associated with descriptions of a wide variety of other\nconcepts. Responses frequently attributed a variable of some kind to the\nstatistic, or a study design detail to the parameter. Implications for\ninstruction and research are discussed, including a call for emphasis on a\nmodeling paradigm in introductory statistics.\n",
"title": "Statistics students' identification of inferential model elements within contexts of their own invention"
}
| null | null |
[
"Statistics"
] | null | true | null |
8908
| null |
Validated
| null | null |
null |
{
"abstract": " We present a conceptually simple, flexible, and general framework for object\ninstance segmentation. Our approach efficiently detects objects in an image\nwhile simultaneously generating a high-quality segmentation mask for each\ninstance. The method, called Mask R-CNN, extends Faster R-CNN by adding a\nbranch for predicting an object mask in parallel with the existing branch for\nbounding box recognition. Mask R-CNN is simple to train and adds only a small\noverhead to Faster R-CNN, running at 5 fps. Moreover, Mask R-CNN is easy to\ngeneralize to other tasks, e.g., allowing us to estimate human poses in the\nsame framework. We show top results in all three tracks of the COCO suite of\nchallenges, including instance segmentation, bounding-box object detection, and\nperson keypoint detection. Without bells and whistles, Mask R-CNN outperforms\nall existing, single-model entries on every task, including the COCO 2016\nchallenge winners. We hope our simple and effective approach will serve as a\nsolid baseline and help ease future research in instance-level recognition.\nCode has been made available at: this https URL\n",
"title": "Mask R-CNN"
}
| null | null |
[
"Computer Science"
] | null | true | null |
8909
| null |
Validated
| null | null |
null |
{
"abstract": " Fishing activities have broad impacts that affect, although not exclusively,\nthe targeted stocks. These impacts affect predators and prey of the harvested\nspecies, as well as the whole ecosystem it inhabits. Ecosystem models can be\nused to study the interactions that occur within a system, including those\nbetween different organisms and those between fisheries and targeted species.\nTrophic web models like Ecopath with Ecosim (EwE) can handle fishing fleets as\na top predator, with top-down impact on harvested organisms. The aim of this\nstudy was to better understand the Icelandic marine ecosystem and the\ninteractions within. This was done by constructing an EwE model of Icelandic\nwaters. The model was run from 1984 to 2013 and was fitted to time series of\nbiomass estimates, landings data and mean annual temperature. The final model\nwas chosen by selecting the model with the lowest Akaike information criterion.\nA skill assessment was performed using the Pearson's correlation coefficient,\nthe coefficient of determination, the modelling efficiency and the reliability\nindex to evaluate the model performance. The model performed satisfactorily\nwhen simulating previously estimated biomass and known landings. Most of the\ngroups with time series were estimated to have top-down control over their\nprey. These are harvested species with direct and/or indirect links to lower\ntrophic levels and future fishing policies should take this into account. This\nmodel could be used as a tool to investigate how such policies could impact the\nmarine ecosystem in Icelandic waters.\n",
"title": "An overview of the marine food web in Icelandic waters using Ecopath with Ecosim"
}
| null | null | null | null | true | null |
8910
| null |
Default
| null | null |
null |
{
"abstract": " Recently, the authors of the present work (together with M. N. Kolountzakis)\nintroduced a new version of the non-commutative Delsarte scheme and applied it\nto the problem of mutually unbiased bases. Here we use this method to\ninvestigate the existence of a finite projective plane of a given order d. In\nparticular, a short new proof is obtained for the nonexistence of a projective\nplane of order 6. For higher orders like 10 and 12, the method is non decisive\nbut could turn out to give important supplementary informations.\n",
"title": "Character tables and the problem of existence of finite projective planes"
}
| null | null | null | null | true | null |
8911
| null |
Default
| null | null |
null |
{
"abstract": " Distributed Generation (DG) units are increasingly installed in the power\nsystems. Distribution Companies (DisCo) can opt to purchase the electricity\nfrom DG in an energy purchase contract to supply the customer demand and reduce\nenergy loss. This paper proposes a framework for optimal contract pricing of\nindependent dispatchable DG units considering competition among them. While DG\nunits tend to increase their profit from the energy purchase contract, DisCo\nminimizes the demand supply cost. Multi-leader follower game theory concept is\nused to analyze the situation in which competing DG units offer the energy\nprice to DisCo and DisCo determines the DG generation. A bi-level approach is\nused to formulate the competition in which each DG problem is the upper-level\nproblem and the DisCo problem is considered as the lower-level one. Combining\nthe optimality conditions ofall upper-level problems with the lower level\nproblem results in a multi-DG equilibrium problem formulated as an equilibrium\nproblem with equilibrium constraints (EPEC). Using a nonlinear approach, the\nEPEC problem is reformulated as a single nonlinear optimization model which is\nsimultaneously solved for all independent DG units. The proposed framework was\napplied to the Modified IEEE 34-Bus Distribution Test System. Performance and\nrobustness of the proposed framework in determining econo-technically fare DG\ncontract price has been demonstrated through a series of analyses.\n",
"title": "A bilevel approach for optimal contract pricing of independent dispatchable DG units in distribution networks"
}
| null | null | null | null | true | null |
8912
| null |
Default
| null | null |
null |
{
"abstract": " Little by little, newspapers are revealing the bright future that Artificial\nIntelligence (AI) is building. Intelligent machines will help everywhere.\nHowever, this bright future has a dark side: a dramatic job market contraction\nbefore its unpredictable transformation. Hence, in a near future, large numbers\nof job seekers will need financial support while catching up with these novel\nunpredictable jobs. This possible job market crisis has an antidote inside. In\nfact, the rise of AI is sustained by the biggest knowledge theft of the recent\nyears. Learning AI machines are extracting knowledge from unaware skilled or\nunskilled workers by analyzing their interactions. By passionately doing their\njobs, these workers are digging their own graves.\nIn this paper, we propose Human-in-the-loop Artificial Intelligence (HIT-AI)\nas a fairer paradigm for Artificial Intelligence systems. HIT-AI will reward\naware and unaware knowledge producers with a different scheme: decisions of AI\nsystems generating revenues will repay the legitimate owners of the knowledge\nused for taking those decisions. As modern Robin Hoods, HIT-AI researchers\nshould fight for a fairer Artificial Intelligence that gives back what it\nsteals.\n",
"title": "Human-in-the-loop Artificial Intelligence"
}
| null | null | null | null | true | null |
8913
| null |
Default
| null | null |
null |
{
"abstract": " The famous \"two-fold cost of sex\" is really the cost of anisogamy -- why\nshould females mate with males who do not contribute resources to offspring,\nrather than isogamous partners who contribute equally? In typical anisogamous\npopulations, a single very fit male can have an enormous number of offspring,\nfar larger than is possible for any female or isogamous individual. If the\nsexual selection on males aligns with the natural selection on females,\nanisogamy thus allows much more rapid adaptation via super-successful males. We\nshow via simulations that this effect can be sufficient to overcome the\ntwo-fold cost and maintain anisogamy against isogamy in populations adapting to\nenvironmental change. The key quantity is the variance in male fitness -- if\nthis exceeds what is possible in an isogamous population, anisogamous\npopulations can win out in direct competition by adapting faster.\n",
"title": "Increased adaptability to rapid environmental change can more than make up for the two-fold cost of males"
}
| null | null | null | null | true | null |
8914
| null |
Default
| null | null |
null |
{
"abstract": " Consider a pair of plane straight-line graphs, whose edges are colored red\nand blue, respectively, and let n be the total complexity of both graphs. We\npresent a O(n log n)-time O(n)-space technique to preprocess such pair of\ngraphs, that enables efficient searches among the red-blue intersections along\nedges of one of the graphs. Our technique has a number of applications to\ngeometric problems. This includes: (1) a solution to the batched red-blue\nsearch problem [Dehne et al. 2006] in O(n log n) queries to the oracle; (2) an\nalgorithm to compute the maximum vertical distance between a pair of 3D\npolyhedral terrains one of which is convex in O(n log n) time, where n is the\ntotal complexity of both terrains; (3) an algorithm to construct the Hausdorff\nVoronoi diagram of a family of point clusters in the plane in O((n+m) log^3 n)\ntime and O(n+m) space, where n is the total number of points in all clusters\nand m is the number of crossings between all clusters; (4) an algorithm to\nconstruct the farthest-color Voronoi diagram of the corners of n axis-aligned\nrectangles in O(n log^2 n) time; (5) an algorithm to solve the stabbing circle\nproblem for n parallel line segments in the plane in optimal O(n log n) time.\nAll these results are new or improve on the best known algorithms.\n",
"title": "Searching edges in the overlap of two plane graphs"
}
| null | null |
[
"Computer Science"
] | null | true | null |
8915
| null |
Validated
| null | null |
null |
{
"abstract": " As more industries integrate machine learning into socially sensitive\ndecision processes like hiring, loan-approval, and parole-granting, we are at\nrisk of perpetuating historical and contemporary socioeconomic disparities.\nThis is a critical problem because on the one hand, organizations who use but\ndo not understand the discriminatory potential of such systems will facilitate\nthe widening of social disparities under the assumption that algorithms are\ncategorically objective. On the other hand, the responsible use of machine\nlearning can help us measure, understand, and mitigate the implicit historical\nbiases in socially sensitive data by expressing implicit decision-making mental\nmodels in terms of explicit statistical models. In this paper we specify,\nimplement, and evaluate a \"fairness-aware\" machine learning interface called\nthemis-ml, which is intended for use by individual data scientists and\nengineers, academic research teams, or larger product teams who use machine\nlearning in production systems.\n",
"title": "Themis-ml: A Fairness-aware Machine Learning Interface for End-to-end Discrimination Discovery and Mitigation"
}
| null | null | null | null | true | null |
8916
| null |
Default
| null | null |
null |
{
"abstract": " This note announces results on the relations between the approach of\nBeilinson and Drinfeld to the geometric Langlands correspondence based on\nconformal field theory, the approach of Kapustin and Witten based on $N=4$ SYM,\nand the AGT-correspondence. The geometric Langlands correspondence is described\nas the Nekrasov-Shatashvili limit of a generalisation of the AGT-correspondence\nin the presence of surface operators. Following the approaches of Kapustin -\nWitten and Nekrasov - Witten we interpret some aspects of the resulting picture\nusing an effective description in terms of two-dimensional sigma models having\nHitchin's moduli spaces as target-manifold.\n",
"title": "Supersymmetric field theories and geometric Langlands: The other side of the coin"
}
| null | null |
[
"Mathematics"
] | null | true | null |
8917
| null |
Validated
| null | null |
null |
{
"abstract": " A new regularisation of the shallow water (and isentropic Euler) equations is\nproposed. The regularised equations are non-dissipative, non-dispersive and\npossess a variational structure. Thus, the mass, the momentum and the energy\nare conserved. Hence, for instance, regularised hydraulic jumps are smooth and\nnon-oscillatory. Another particularly interesting feature of this\nregularisation is that smoothed `shocks' propagates at exactly the same speed\nas the original discontinuous ones. The performance of the new model is\nillustrated numerically on some dam-break test cases, which are classical in\nthe hyperbolic realm.\n",
"title": "Non-dispersive conservative regularisation of nonlinear shallow water (and isothermal Euler) equations"
}
| null | null |
[
"Physics"
] | null | true | null |
8918
| null |
Validated
| null | null |
null |
{
"abstract": " Recently, Andrews, Dixit and Yee defined two partition functions\n$p_{\\omega}(n)$ and $p_{\\nu}(n)$ that are related with Ramanujan's mock theta\nfunctions $\\omega(q)$ and $\\nu(q)$, respectively. In this paper, we present two\nvariable generalizations of their results. As an application, we reprove their\nresults on $p_{\\omega}(n)$ and $p_{\\nu}(n)$ that are analogous to Euler's\npentagonal number theorem.\n",
"title": "Some Identities associated with mock theta functions $ω(q)$ and $ν(q)$"
}
| null | null | null | null | true | null |
8919
| null |
Default
| null | null |
null |
{
"abstract": " We examine nonlinear dynamical systems of ordinary differential equations or\ndifferential algebraic equations. In an uncertainty quantification, physical\nparameters are replaced by random variables. The inner variables as well as a\nquantity of interest are expanded into series with orthogonal basis functions\nlike the polynomial chaos expansions, for example. On the one hand, the\nstochastic Galerkin method yields a large coupled dynamical system. On the\nother hand, a stochastic collocation method, which uses a quadrature rule or a\nsampling scheme, can be written in the form of a large weakly coupled dynamical\nsystem. We apply projection-based methods of nonlinear model order reduction to\nthe large systems. A reduced-order model implies a low-dimensional\nrepresentation of the quantity of interest. We focus on model order reduction\nby proper orthogonal decomposition. The error of a best approximation located\nin a low-dimensional subspace is analysed. We illustrate results of numerical\ncomputations for test examples.\n",
"title": "Model order reduction for random nonlinear dynamical systems and low-dimensional representations for their quantities of interest"
}
| null | null | null | null | true | null |
8920
| null |
Default
| null | null |
null |
{
"abstract": " We provide an algorithm that computes a set of generators for any complete\nideal in a smooth complex surface. More interestingly, these generators admit a\npresentation as monomials in a set of maximal contact elements associated to\nthe minimal log-resolution of the ideal. Furthermore, the monomial expression\ngiven by our method is an equisingularity invariant of the ideal. As an\noutcome, we provide a geometric method to compute the integral closure of a\nplanar ideal and we apply our algorithm to some families of complete ideals.\n",
"title": "Monomial generators of complete planar ideals"
}
| null | null | null | null | true | null |
8921
| null |
Default
| null | null |
null |
{
"abstract": " Over the last decade, digital media (web or app publishers) generalized the\nuse of real time ad auctions to sell their ad spaces. Multiple auction\nplatforms, also called Supply-Side Platforms (SSP), were created. Because of\nthis multiplicity, publishers started to create competition between SSPs. In\nthis setting, there are two successive auctions: a second price auction in each\nSSP and a secondary, first price auction, called header bidding auction,\nbetween SSPs.In this paper, we consider an SSP competing with other SSPs for ad\nspaces. The SSP acts as an intermediary between an advertiser wanting to buy ad\nspaces and a web publisher wanting to sell its ad spaces, and needs to define a\nbidding strategy to be able to deliver to the advertisers as many ads as\npossible while spending as little as possible. The revenue optimization of this\nSSP can be written as a contextual bandit problem, where the context consists\nof the information available about the ad opportunity, such as properties of\nthe internet user or of the ad placement.Using classical multi-armed bandit\nstrategies (such as the original versions of UCB and EXP3) is inefficient in\nthis setting and yields a low convergence speed, as the arms are very\ncorrelated. In this paper we design and experiment a version of the Thompson\nSampling algorithm that easily takes this correlation into account. We combine\nthis bayesian algorithm with a particle filter, which permits to handle\nnon-stationarity by sequentially estimating the distribution of the highest bid\nto beat in order to win an auction. We apply this methodology on two real\nauction datasets, and show that it significantly outperforms more classical\napproaches.The strategy defined in this paper is being developed to be deployed\non thousands of publishers worldwide.\n",
"title": "Optimization of a SSP's Header Bidding Strategy using Thompson Sampling"
}
| null | null |
[
"Statistics"
] | null | true | null |
8922
| null |
Validated
| null | null |
null |
{
"abstract": " Dyonic 1/4-BPS states in Type IIB string theory compactified on $\\mathrm{K}3\n\\times T^2$ are counted by meromorphic Jacobi forms. The finite parts of these\nfunctions, which are mixed mock Jacobi forms, account for the degeneracy of\nstates stable throughout the moduli space of the compactification. In this\npaper, we obtain an exact asymptotic expansion for their Fourier coefficients,\nrefining the Hardy-Ramanujan-Littlewood circle method to deal with their\nmixed-mock character. The result is compared to a low-energy supergravity\ncomputation of the exact entropy of extremal dyonic 1/4-BPS single-centered\nblack holes, obtained by applying supersymmetric localization techniques to the\nquantum entropy function.\n",
"title": "Mixed Rademacher and BPS Black Holes"
}
| null | null | null | null | true | null |
8923
| null |
Default
| null | null |
null |
{
"abstract": " We introduce a family of tensor network states that we term semi-injective\nProjected Entangled-Pair States (PEPS). They extend the class of injective PEPS\nand include other states, like the ground states of the AKLT and the CZX models\nin square lattices. We construct parent Hamiltonians for which semi-injective\nPEPS are unique ground states. We also determine the necessary and sufficient\nconditions for two tensors to generate the same family of such states in two\nspatial dimensions. Using this result, we show that the third cohomology\nlabeling of Symmetry Protected Topological phases extends to semi-injective\nPEPS.\n",
"title": "A generalization of the injectivity condition for Projected Entangled Pair States"
}
| null | null | null | null | true | null |
8924
| null |
Default
| null | null |
null |
{
"abstract": " Single atoms form a model system for understanding the limits of single\nphoton detection. Here, we develop a non-Markov theory of single-photon\nabsorption by a two-level atom to place limits on the absorption (transduction)\ntime. We show the existence of a finite rise time in the probability of\nexcitation of the atom during the absorption event which is infinitely fast in\nprevious Markov theories. This rise time is governed by the bandwidth of the\natom-field interaction spectrum and leads to a fundamental jitter in\ntime-stamping the absorption event. Our theoretical framework captures both the\nweak and strong atom-field coupling regimes and sheds light on the spectral\nmatching between the interaction bandwidth and single photon Fock state pulse\nspectrum. Our work opens questions whether such jitter in the absorption event\ncan be observed in a multi-mode realistic single photon detector. Finally, we\nalso shed light on the fundamental differences between linear and nonlinear\ndetector outputs for single photon Fock state vs. coherent state pulses.\n",
"title": "Limits to single photon transduction by a single atom: Non-Markov theory"
}
| null | null | null | null | true | null |
8925
| null |
Default
| null | null |
null |
{
"abstract": " A verbal autopsy (VA) consists of a survey with a relative or close contact\nof a person who has recently died. VA surveys are commonly used to infer likely\ncauses of death for individuals when deaths happen outside of hospitals or\nhealthcare facilities. Several statistical and algorithmic methods are\navailable to assign cause of death using VA surveys. Each of these methods\nrequire as inputs some information about the joint distribution of symptoms and\ncauses. In this note, we examine the generalizability of this symptom-cause\ninformation by comparing different automated coding methods using various\ncombinations of inputs and evaluation data. VA algorithm performance is\naffected by both the specific SCI themselves and the logic of a given\nalgorithm. Using a variety of performance metrics for all existing VA\nalgorithms, we demonstrate that in general the adequacy of the information\nabout the joint distribution between symptoms and cause affects performance at\nleast as much or more than algorithm logic.\n",
"title": "Quantifying the Contributions of Training Data and Algorithm Logic to the Performance of Automated Cause-assignment Algorithms for Verbal Autopsy"
}
| null | null | null | null | true | null |
8926
| null |
Default
| null | null |
null |
{
"abstract": " We report on a versatile, highly controllable hybrid cold Rydberg atom fiber\ninterface, based on laser cooled atoms transported into a hollow core\nKagomé crystal fiber. Our experiments are the first to demonstrate the\nfeasibility of exciting cold Rydberg atoms inside a hollow core fiber and we\nstudy the influence of the fiber on Rydberg electromagnetically induced\ntransparency (EIT) signals. Using a temporally resolved detection method to\ndistinguish between excitation and loss, we observe two different regimes of\nthe Rydberg excitations: one EIT regime and one regime dominated by atom loss.\nThese results are a substantial advancement towards future use of our system\nfor quantum simulation or information.\n",
"title": "Rydberg excitation of cold atoms inside a hollow core fiber"
}
| null | null | null | null | true | null |
8927
| null |
Default
| null | null |
null |
{
"abstract": " In some planetary systems the orbital periods of two of its members present a\ncommensurability, usually known by mean-motion resonance. These resonances\ngreatly enhance the mutual gravitational influence of the planets. As a\nconsequence, these systems present uncommon behaviours and their motions need\nto be studied with specific methods. Some features are unique and allow us a\nbetter understanding and characterisation of these systems. Moreover,\nmean-motion resonances are a result of an early migration of the orbits in an\naccretion disk, so it is possible to derive constraints on their formation.\nHere we review the dynamics of a pair of resonant planets and explain how their\norbits evolve in time. We apply our results to the HD45365 planetary system\n",
"title": "Dynamics and evolution of planets in mean-motion resonances"
}
| null | null |
[
"Physics",
"Mathematics"
] | null | true | null |
8928
| null |
Validated
| null | null |
null |
{
"abstract": " We consider the motion of a nonrelativistic electron in the field of two\nstrong monochromatic light waves propagating counter to each other. The matrix\nelements of emission and absorption are found. An expression is obtained for\nthe gain of a weak test wave by using such matrix elements.\n",
"title": "The Gain in the Field of Two Electromagnetic Waves"
}
| null | null | null | null | true | null |
8929
| null |
Default
| null | null |
null |
{
"abstract": " There is a renewed interest in weak model sets due to their connection to\n$\\mathcal B$-free systems, which emerged from Sarnak's program on the Möbius\ndisjointness conjecture. Here we continue our recent investigation\n[arXiv:1511.06137] of the extended hull ${\\mathcal M}^{\\scriptscriptstyle\nG}_{\\scriptscriptstyle W}$, a dynamical system naturally associated to a weak\nmodel set in an abelian group $G$ with relatively compact window $W$. For\nwindows having a nowhere dense boundary (this includes compact windows), we\nidentify the maximal equicontinuous factor of ${\\mathcal M}^{\\scriptscriptstyle\nG}_{\\scriptscriptstyle W}$ and give a sufficient condition when ${\\mathcal\nM}^{\\scriptscriptstyle G}_{\\scriptscriptstyle W}$ is an almost 1:1 extension of\nits maximal equicontinuous factor. If the window is measurable with positive\nHaar measure and is almost compact, then the system ${\\mathcal\nM}^{\\scriptscriptstyle G}_{\\scriptscriptstyle W}$ equipped with its Mirsky\nmeasure is isomorphic to its Kronecker factor. For general nontrivial ergodic\nprobability measures on ${\\mathcal M}^{\\scriptscriptstyle\nG}_{\\scriptscriptstyle W}$, we provide a kind of lower bound for the Kronecker\nfactor. All relevant factor systems are natural $G$-actions on quotient\nsubgroups of the torus underlying the weak model set. These are obtained by\nfactoring out suitable window periods. Our results are specialised to the usual\nhull of the weak model set, and they are also interpreted for ${\\mathcal\nB}$-free systems.\n",
"title": "Periods and factors of weak model sets"
}
| null | null | null | null | true | null |
8930
| null |
Default
| null | null |
null |
{
"abstract": " Citation metrics are analytic measures used to evaluate the usage, impact and\ndissemination of scientific research. Traditionally, citation metrics have been\nindependently measured at each level of the publication pyramid, namely at the\narticle-level, at the author-level, and at the journal-level. The most commonly\nused metrics have been focused on journal-level measurements, such as the\nImpact Factor and the Eigenfactor, as well as on researcher-level metrics like\nthe Hirsch index (h-index) and i10 index. On the other hand, reliable\narticle-level metrics are less widespread, and are often reserved to\nnon-standardized and non-scientific characteristics of individual articles,\nsuch as views, citations, downloads, and mentions in social and news media.\nThese characteristics are known as 'altmetrics'. However, when the number of\nviews and citations are similar between two articles, no discriminating measure\ncurrently exists with which to assess and compare each articles' individual\nimpact. Given the modern, exponentially growing scientific literature,\nscientists and readers of Science need optimized, reliable, objective methods\nfor managing, measuring and comparing research outputs and individual\npublications. To this end, I hereby describe and propose a new standardized\narticle-level metric henceforth known as the 'Individual Impact Index\nStatistic', or $i^3$ for short. The $i^3$ is a weighted algorithm that takes\nadvantage of the peer-review process, and considers a number of characteristics\nof individual scientific publications in order to yield a standardized and\nreadily comparable measure of impact and dissemination. The strengths,\nlimitations, and potential uses of this novel metric are also discussed.\n",
"title": "The Individual Impact Index ($i^3$) Statistic: A Novel Article-Level Citation Metric"
}
| null | null | null | null | true | null |
8931
| null |
Default
| null | null |
null |
{
"abstract": " The physical properties of polycrystalline materials depend on their\nmicrostructure, which is the nano-to-centimeter-scale arrangement of phases and\ndefects in their interior. Such microstructure depends on the shape,\ncrystallographic phase and orientation, and interfacing of the grains\nconstituting the material. This article presents a new non-destructive 3D\ntechnique to study bulk samples with sizes in the cm range with a resolution of\nhundred micrometers: time-of-flight three-dimensional neutron diffraction (ToF\n3DND). Compared to existing analogous X-ray diffraction techniques, ToF 3DND\nenables studies of samples that can be both larger in size and made of heavier\nelements. Moreover, ToF 3DND facilitates the use of complicated sample\nenvironments. The basic ToF 3DND setup, utilizing an imaging detector with high\nspatial and temporal resolution, can easily be implemented at a time-of-flight\nneutron beamline. The technique was developed and tested with data collected at\nthe Materials and Life Science Experimental Facility of the Japan Proton\nAccelerator Complex (J-PARC) for an iron sample. We successfully reconstructed\nthe shape of 108 grains and developed an indexing procedure. The reconstruction\nalgorithms have been validated by reconstructing two stacked Co-Ni-Ga single\ncrystals and by comparison with a grain map obtained by post-mortem electron\nbackscatter diffraction (EBSD).\n",
"title": "Time-of-Flight Three Dimensional Neutron Diffraction in Transmission Mode for Mapping Crystal Grain Structures"
}
| null | null | null | null | true | null |
8932
| null |
Default
| null | null |
null |
{
"abstract": " We introduce and study new categories T(g,k)of integrable sl(\\infty)-modules\nwhich depend on the choice of a certain reductive subalgebra k in g=sl(\\infty).\nThe simple objects of these categories are tensor modules as in the previously\nstudied category, however, the choice of k provides more flexibility of\nnonsimple modules. We then choose k to have two infinite-dimensional diagonal\nblocks, and show that a certain injective object K(m|n) in T(g,k) realizes a\ncategorical sl(\\infty)-action on the integral category O(m|n) of the Lie\nsuperalgebra gl(m|n). We show that the socle of K(m|n) is generated by the\nprojective modules in O(m|n), and compute the socle filtration of K(m|n)\nexplicitly. We conjecture that the socle filtration of K(m|n) reflects a\n\"degree of atypicality filtration\" on the category O(m|n). We also conjecture\nthat a natural tensor filtration on K(m|n) arises via the Duflo--Serganova\nfunctor sending the category O(m|n) to O(m-1|n-1). We prove this latter\nconjecture for a direct summand of K(m|n) corresponding to the\nfinite-dimensional gl(m|n)-modules.\n",
"title": "Integrable $sl(\\infty)$-modules and Category $\\mathcal O$ for $\\mathfrak{gl}(m|n)$"
}
| null | null | null | null | true | null |
8933
| null |
Default
| null | null |
null |
{
"abstract": " Using in situ grazing-incidence x-ray scattering, we have measured the\ndiffuse scattering from islands that form during layer-by-layer growth of GaN\nby metal-organic vapor phase epitaxy on the (1010) m-plane surface. The diffuse\nscattering is extended in the (0001) in-plane direction in reciprocal space,\nindicating a strong anisotropy with islands elongated along [1 $\\overline{2}$\n10] and closely spaced along [0001]. This is confirmed by atomic force\nmicroscopy of a quenched sample. Islands were characterized as a function of\ngrowth rate G and temperature. The island spacing along [0001] observed during\nthe growth of the first monolayer obeys a power-law dependence on growth rate\nG$^{-n}$, with an exponent $n = 0.25 \\pm 0.02$. Results are in agreement with\nrecent kinetic Monte Carlo simulations, indicating that elongated islands\nresult from the dominant anisotropy in step edge energy and not from surface\ndiffusion anisotropy. The observed power-law exponent can be explained using a\nsimple steady-state model, which gives n = 1/4.\n",
"title": "Island dynamics and anisotropy during vapor phase epitaxy of m-plane GaN"
}
| null | null | null | null | true | null |
8934
| null |
Default
| null | null |
null |
{
"abstract": " The well-known Komlós-Major-Tusnády inequalities [Z. Wahrsch. Verw.\nGebiete 32 (1975) 111-131; Z. Wahrsch. Verw. Gebiete 34 (1976) 33-58] provide\nsharp inequalities to partial sums of iid standard exponential random variables\nby a sequence of standard Brownian motions. In this paper, we employ these\nresults to establish Gaussian approximations to weighted increments of uniform\nempirical and quantile processes. This approach provides rates to the\napproximations which, among others, have direct applications to statistics of\nextreme values for randomly censored data.\n",
"title": "Komlós-Major-Tusnády approximations to increments of uniform empirical processes"
}
| null | null | null | null | true | null |
8935
| null |
Default
| null | null |
null |
{
"abstract": " We consider the problem of learning the level set for which a noisy black-box\nfunction exceeds a given threshold.\nTo efficiently reconstruct the level set, we investigate Gaussian process\n(GP) metamodels. Our focus is on strongly stochastic samplers, in particular\nwith heavy-tailed simulation noise and low signal-to-noise ratio.\nTo guard against noise misspecification, we assess the performance of three\nvariants: (i) GPs with Student-$t$ observations; (ii) Student-$t$ processes\n(TPs); and (iii) classification GPs modeling the sign of the response. As a\nfourth extension, we study GP surrogates with monotonicity constraints that are\nrelevant when the level set is known to be connected. In conjunction with these\nmetamodels, we analyze several acquisition functions for guiding the sequential\nexperimental designs, extending existing stepwise uncertainty reduction\ncriteria to the stochastic contour-finding context. This also motivates our\ndevelopment of (approximate) updating formulas to efficiently compute such\nacquisition functions. Our schemes are benchmarked by using a variety of\nsynthetic experiments in 1--6 dimensions. We also consider an application of\nlevel set estimation for determining the optimal exercise policy and valuation\nof Bermudan options in finance.\n",
"title": "Evaluating Gaussian Process Metamodels and Sequential Designs for Noisy Level Set Estimation"
}
| null | null |
[
"Statistics"
] | null | true | null |
8936
| null |
Validated
| null | null |
null |
{
"abstract": " In the last decade, many business applications have moved into the cloud. In\nparticular, the \"database-as-a-service\" paradigm has become mainstream. While\nexisting multi-tenant data management systems focus on single-tenant query\nprocessing, we believe that it is time to rethink how queries can be processed\nacross multiple tenants in such a way that we do not only gain more valuable\ninsights, but also at minimal cost. As we will argue in this paper, standard\nSQL semantics are insufficient to process cross-tenant queries in an\nunambiguous way, which is why existing systems use other, expensive means like\nETL or data integration. We first propose MTSQL, a set of extensions to\nstandard SQL, which fixes the ambiguity problem. Next, we present MTBase, a\nquery processing middleware that efficiently processes MTSQL on top of SQL. As\nwe will see, there is a canonical, provably correct, rewrite algorithm from\nMTSQL to SQL, which may however result in poor query execution performance,\neven on high-performance database products. We further show that with\ncarefully-designed optimizations, execution times can be reduced in such ways\nthat the difference to single-tenant queries becomes marginal.\n",
"title": "MTBase: Optimizing Cross-Tenant Database Queries"
}
| null | null |
[
"Computer Science"
] | null | true | null |
8937
| null |
Validated
| null | null |
null |
{
"abstract": " We introduce KiNetX, a fully automated meta-algorithm for the kinetic\nanalysis of complex chemical reaction networks derived from semi-accurate but\nefficient electronic structure calculations. It is designed to (i) accelerate\nthe automated exploration of such networks, and (ii) cope with model-inherent\nerrors in electronic structure calculations on elementary reaction steps. We\ndeveloped and implemented KiNetX to possess three features. First, KiNetX\nevaluates the kinetic relevance of every species in a (yet incomplete) reaction\nnetwork to confine the search for new elementary reaction steps only to those\nspecies that are considered possibly relevant. Second, KiNetX identifies and\neliminates all kinetically irrelevant species and elementary reactions to\nreduce a complex network graph to a comprehensible mechanism. Third, KiNetX\nestimates the sensitivity of species concentrations toward changes in\nindividual rate constants (derived from relative free energies), which allows\nus to systematically select the most efficient electronic structure model for\neach elementary reaction given a predefined accuracy. The novelty of KiNetX\nconsists in the rigorous propagation of correlated free-energy uncertainty\nthrough all steps of our kinetic analyis. To examine the performance of KiNetX,\nwe developed AutoNetGen. It semirandomly generates chemistry-mimicking reaction\nnetworks by encoding chemical logic into their underlying graph structure.\nAutoNetGen allows us to consider a vast number of distinct chemistry-like\nscenarios and, hence, to discuss assess the importance of rigorous uncertainty\npropagation in a statistical context. Our results reveal that KiNetX reliably\nsupports the deduction of product ratios, dominant reaction pathways, and\npossibly other network properties from semi-accurate electronic structure data.\n",
"title": "Mechanism Deduction from Noisy Chemical Reaction Networks"
}
| null | null | null | null | true | null |
8938
| null |
Default
| null | null |
null |
{
"abstract": " In the present contribution, we study the Landau-Lifshitz-Gilbert equation\nwith two versions of structural derivatives recently proposed: the scale\n$q-$derivative in the non-extensive statistical mechanics and the axiomatic\nmetric derivative, which presents Mittag-Leffler functions as eigenfunctions.\nThe use of structural derivatives aims to take into account long-range forces,\npossible non-manifest or hidden interactions and the dimensionality of space.\nHaving this purpose in mind, we build up an evolution operator and a deformed\nversion of the LLG equation. Damping in the oscillations naturally show up\nwithout an explicit Gilbert damping term.\n",
"title": "Structural scale $q-$derivative and the LLG-Equation in a scenario with fractionality"
}
| null | null | null | null | true | null |
8939
| null |
Default
| null | null |
null |
{
"abstract": " A new approach to perform analog optical differentiation is presented using\nhalf-wavelength slabs. First, a half-wavelength dielectric slab is used to\ndesign a first order differentiator. The latter works properly for both major\npolarizations, in contrast to designs based on Brewster effect [Opt. Lett. 41,\n3467 (2016)]. Inspired by the proposed dielectric differentiator, and by\nexploiting the unique features of graphene, we further design and demonstrate a\nreconfigurable and highly miniaturized differentiator using a half-wavelength\nplasmonic graphene film. To the best of our knowledge, our proposed\ngraphene-based differentiator is even smaller than the most compact\ndifferentiator presented so far [Opt. Lett. 40, 5239 (2015)].\n",
"title": "Analog Optical Computing by Half-Wavelength Slabs"
}
| null | null | null | null | true | null |
8940
| null |
Default
| null | null |
null |
{
"abstract": " We review the problem of defining and inferring a \"state\" for a control\nsystem based on complex, high-dimensional, highly uncertain measurement streams\nsuch as videos. Such a state, or representation, should contain all and only\nthe information needed for control, and discount nuisance variability in the\ndata. It should also have finite complexity, ideally modulated depending on\navailable resources. This representation is what we want to store in memory in\nlieu of the data, as it \"separates\" the control task from the measurement\nprocess. For the trivial case with no dynamics, a representation can be\ninferred by minimizing the Information Bottleneck Lagrangian in a function\nclass realized by deep neural networks. The resulting representation has much\nhigher dimension than the data, already in the millions, but it is smaller in\nthe sense of information content, retaining only what is needed for the task.\nThis process also yields representations that are invariant to nuisance factors\nand having maximally independent components. We extend these ideas to the\ndynamic case, where the representation is the posterior density of the task\nvariable given the measurements up to the current time, which is in general\nmuch simpler than the prediction density maintained by the classical Bayesian\nfilter. Again this can be finitely-parametrized using a deep neural network,\nand already some applications are beginning to emerge. No explicit assumption\nof Markovianity is needed; instead, complexity trades off approximation of an\noptimal representation, including the degree of Markovianity.\n",
"title": "A Separation Principle for Control in the Age of Deep Learning"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
8941
| null |
Validated
| null | null |
null |
{
"abstract": " Quantization can improve the execution latency and energy efficiency of\nneural networks on both commodity GPUs and specialized accelerators. The\nmajority of existing literature focuses on training quantized DNNs, while this\nwork examines the less-studied topic of quantizing a floating-point model\nwithout (re)training. DNN weights and activations follow a bell-shaped\ndistribution post-training, while practical hardware uses a linear quantization\ngrid. This leads to challenges in dealing with outliers in the distribution.\nPrior work has addressed this by clipping the outliers or using specialized\nhardware. In this work, we propose outlier channel splitting (OCS), which\nduplicates channels containing outliers, then halves the channel values. The\nnetwork remains functionally identical, but affected outliers are moved toward\nthe center of the distribution. OCS requires no additional training and works\non commodity hardware. Experimental evaluation on ImageNet classification and\nlanguage modeling shows that OCS can outperform state-of-the-art clipping\ntechniques with only minor overhead.\n",
"title": "Improving Neural Network Quantization using Outlier Channel Splitting"
}
| null | null | null | null | true | null |
8942
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we provide some new results for the Weibull-R family of\ndistributions (Alzaghal, Ghosh and Alzaatreh (2016)). We derive some new\nstructural properties of the Weibull-R family of distributions. We provide\nvarious characterizations of the family via conditional moments, some functions\nof order statistics and via record values.\n",
"title": "On some further properties and application of Weibull-R family of distributions"
}
| null | null | null | null | true | null |
8943
| null |
Default
| null | null |
null |
{
"abstract": " The principle of material frame indifference is shown to be incompatible with\nthe basic balance laws of continuum mechanics. In its role of providing\nconstraints on possible constitutive prescriptions it must be replaced by the\nclassical principle of Galilean invariance.\n",
"title": "Rejection of the principle of material frame indifference"
}
| null | null | null | null | true | null |
8944
| null |
Default
| null | null |
null |
{
"abstract": " Recently, we introduced the notion of flow (depending on time) of\nfinite-dimensional algebras. A flow of algebras (FA) is a particular case of a\ncontinuous-time dynamical system whose states are finite-dimensional algebras\nwith (cubic) matrices of structural constants satisfying an analogue of the\nKolmogorov-Chapman equation (KCE). Since there are several kinds of\nmultiplications between cubic matrices one has fix a multiplication first and\nthen consider the KCE with respect to the fixed multiplication. The existence\nof a solution for the KCE provides the existence of an FA. In this paper our\naim is to find sufficient conditions on the multiplications under which the\ncorresponding KCE has a solution. Mainly our conditions are given on the\nalgebra of cubic matrices (ACM) considered with respect to a fixed\nmultiplication of cubic matrices. Under some assumptions on the ACM (e.g. power\nassociative, unital, associative, commutative) we describe a wide class of FAs,\nwhich contain algebras of arbitrary finite dimension. In particular, adapting\nthe theory of continuous-time Markov processes, we construct a class of FAs\ngiven by the matrix exponent of cubic matrices. Moreover, we remarkably extend\nthe set of FAs given with respect to the Maksimov's multiplications of our\nprevious paper (J. Algebra 470 (2017) 263--288). For several FAs we study the\ntime-dependent behavior (dynamics) of the algebras. We derive a system of\ndifferential equations for FAs.\n",
"title": "Construction of flows of finite-dimensional algebras"
}
| null | null | null | null | true | null |
8945
| null |
Default
| null | null |
null |
{
"abstract": " Aims. We present new IRAM Plateau de Bure Interferometer observations of Arp\n220 in HCN, HCO$^{+}$, HN$^{13}$C J=1-0, C$_{2}$H N=1-0, SiO J = 2-1, HNCO\nJ$_{k,k'}$ = 5$_{0,4}$ - 4$_{0,4}$, CH$_{3}$CN(6-5), CS J=2-1 and 5-4 and\n$^{13}$CO J=1-0 and 2-1 and of NGC 6240 in HCN, HCO$^{+}$ J = 1-0 and C$_{2}$H\nN = 1-0. In addition, we present Atacama Large Millimeter/submillmeter Array\nscience verification observations of Arp 220 in CS J = 4-3 and\nCH$_{3}$CN(10-9). Various lines are used to analyse the physical conditions of\nthe molecular gas including the [$^{12}$CO]/[$^{13}$CO] and\n[$^{12}$CO]/[C$^{18}$O] abundance ratios. These observations will be made\navailable to the public. Methods. We create brightness temperature line ratio\nmaps to present the different physical conditions across Arp 220 and NGC 6240.\nIn addition, we use the radiative transfer code RADEX and a Monte Carlo Markov\nChain likelihood code to model the $^{12}$CO, $^{13}$CO and C$^{18}$O lines of\nArp 220 at ~2\" (~700 pc) scales, where the $^{12}$CO and C$^{18}$O measurements\nwere obtained from literature. Results. Line ratios of optically thick lines\nsuch as $^{12}$CO show smoothly varying ratios while the line ratios of\noptically thin lines such as $^{13}$CO show a east-west gradient across Arp\n220. The HCN/HCO$^{+}$ line ratio differs between Arp 220 and NGC 6240, where\nArp 220 has line ratios above 2 and NGC 6240 below 1. The radiative transfer\nanalysis solution is consistent with a warm (~40 K), moderately dense\n(~10$^{3.4}$ cm$^{-3}$) molecular gas component averaged over the two nuclei.\nWe find [$^{12}$CO]/[$^{13}$CO] and [$^{12}$CO]/[C$^{18}$O] abundance ratios of\n~90 for both. The abundance enhancement of C$^{18}$O can be explained by\nstellar nucleosynthesis enrichment of the interstellar medium.\n",
"title": "PdBI U/LIRG Survey (PULS): Dense Molecular Gas in Arp 220 and NGC 6240"
}
| null | null | null | null | true | null |
8946
| null |
Default
| null | null |
null |
{
"abstract": " In recent years, mobile devices (e.g., smartphones and tablets) have met an\nincreasing commercial success and have become a fundamental element of the\neveryday life for billions of people all around the world. Mobile devices are\nused not only for traditional communication activities (e.g., voice calls and\nmessages) but also for more advanced tasks made possible by an enormous amount\nof multi-purpose applications (e.g., finance, gaming, and shopping). As a\nresult, those devices generate a significant network traffic (a consistent part\nof the overall Internet traffic). For this reason, the research community has\nbeen investigating security and privacy issues that are related to the network\ntraffic generated by mobile devices, which could be analyzed to obtain\ninformation useful for a variety of goals (ranging from device security and\nnetwork optimization, to fine-grained user profiling).\nIn this paper, we review the works that contributed to the state of the art\nof network traffic analysis targeting mobile devices. In particular, we present\na systematic classification of the works in the literature according to three\ncriteria: (i) the goal of the analysis; (ii) the point where the network\ntraffic is captured; and (iii) the targeted mobile platforms. In this survey,\nwe consider points of capturing such as Wi-Fi Access Points, software\nsimulation, and inside real mobile devices or emulators. For the surveyed\nworks, we review and compare analysis techniques, validation methods, and\nachieved results. We also discuss possible countermeasures, challenges and\npossible directions for future research on mobile traffic analysis and other\nemerging domains (e.g., Internet of Things). We believe our survey will be a\nreference work for researchers and practitioners in this research field.\n",
"title": "The Dark Side(-Channel) of Mobile Devices: A Survey on Network Traffic Analysis"
}
| null | null | null | null | true | null |
8947
| null |
Default
| null | null |
null |
{
"abstract": " We consider the Rosenzweig-Porter model $H = V + \\sqrt{T}\\, \\Phi$, where $V$\nis a $N \\times N$ diagonal matrix, $\\Phi$ is drawn from the $N \\times N$\nGaussian Orthogonal Ensemble, and $N^{-1} \\ll T \\ll 1$. We prove that the\neigenfunctions of $H$ are typically supported in a set of approximately $NT$\nsites, thereby confirming the existence of a previously conjectured non-ergodic\ndelocalized phase. Our proof is based on martingale estimates along the\ncharacteristic curves of the stochastic advection equation satisfied by the\nlocal resolvent of the Brownian motion representation of $H$.\n",
"title": "Non-Ergodic Delocalization in the Rosenzweig-Porter Model"
}
| null | null | null | null | true | null |
8948
| null |
Default
| null | null |
null |
{
"abstract": " In a seminal paper, McAfee (1992) presented a truthful mechanism for double\nauctions, attaining asymptotically-optimal gain-from-trade without any prior\ninformation on the valuations of the traders. McAfee's mechanism handles\nsingle-parametric agents, allowing each seller to sell a single unit and each\nbuyer to buy a single unit. This paper presents a double-auction mechanism that\nhandles multi-parametric agents and allows multiple units per trader, as long\nas the valuation functions of all traders have decreasing marginal returns. The\nmechanism is prior-free, ex-post individually-rational, dominant-strategy\ntruthful and strongly-budget-balanced. Its gain-from-trade approaches the\noptimum when the market size is sufficiently large.\n",
"title": "MUDA: A Truthful Multi-Unit Double-Auction Mechanism"
}
| null | null | null | null | true | null |
8949
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we propose multi-variable LSTM capable of accurate forecasting\nand variable importance interpretation for time series with exogenous\nvariables. Current attention mechanism in recurrent neural networks mostly\nfocuses on the temporal aspect of data and falls short of characterizing\nvariable importance. To this end, the multi-variable LSTM equipped with\ntensorized hidden states is developed to learn hidden states for individual\nvariables, which give rise to our mixture temporal and variable attention.\nBased on such attention mechanism, we infer and quantify variable importance.\nExtensive experiments using real datasets with Granger-causality test and the\nsynthetic dataset with ground truth demonstrate the prediction performance and\ninterpretability of multi-variable LSTM in comparison to a variety of\nbaselines. It exhibits the prospect of multi-variable LSTM as an end-to-end\nframework for both forecasting and knowledge discovery.\n",
"title": "Multi-variable LSTM neural network for autoregressive exogenous model"
}
| null | null |
[
"Statistics"
] | null | true | null |
8950
| null |
Validated
| null | null |
null |
{
"abstract": " Spectral images captured by satellites and radio-telescopes are analyzed to\nobtain information about geological compositions distributions, distant asters\nas well as undersea terrain. Spectral images usually contain tens to hundreds\nof continuous narrow spectral bands and are widely used in various fields. But\nthe vast majority of those image signals are beyond the visible range, which\ncalls for special visualization technique. The visualizations of spectral\nimages shall convey as much information as possible from the original signal\nand facilitate image interpretation. However, most of the existing visualizatio\nmethods display spectral images in false colors, which contradict with human's\nexperience and expectation. In this paper, we present a novel visualization\ngenerative adversarial network (GAN) to display spectral images in natural\ncolors. To achieve our goal, we propose a loss function which consists of an\nadversarial loss and a structure loss. The adversarial loss pushes our solution\nto the natural image distribution using a discriminator network that is trained\nto differentiate between false-color images and natural-color images. We also\nuse a cycle loss as the structure constraint to guarantee structure\nconsistency. Experimental results show that our method is able to generate\nstructure-preserved and natural-looking visualizations.\n",
"title": "Spectral Image Visualization Using Generative Adversarial Networks"
}
| null | null | null | null | true | null |
8951
| null |
Default
| null | null |
null |
{
"abstract": " We introduce simulations aimed at assessing how well weak gravitational\nlensing of 21cm radiation from the Epoch of Reionization ($z \\sim 8$) can be\nmeasured by an SKA-like radio telescope. A simulation pipeline has been\nimplemented to study the performance of lensing reconstruction techniques. We\nshow how well the lensing signal can be reconstructed using the\nthree-dimensional quadratic lensing estimator in Fourier space assuming\ndifferent survey strategies. The numerical code introduced in this work is\ncapable of dealing with issues that can not be treated analytically such as the\ndiscreteness of visibility measurements and the inclusion of a realistic model\nfor the antennae distribution. This paves the way for future numerical studies\nimplementing more realistic reionization models, foreground subtraction\nschemes, and testing the performance of lensing estimators that take into\naccount the non-Gaussian distribution of HI after reionization. If multiple\nfrequency channels covering $z \\sim 7-11.6$ are combined, Phase 1 of SKA-Low\nshould be able to obtain good quality images of the lensing potential with a\ntotal resolution of $\\sim 1.6$ arcmin. The SKA-Low Phase 2 should be capable of\nproviding images with high-fidelity even using data from $z\\sim 7.7 - 8.3$. We\nperform tests aimed at evaluating the numerical implementation of the mapping\nreconstruction. We also discuss the possibility of measuring an accurate\nlensing power spectrum. Combining data from $z \\sim 7-11.6$ using the SKA2-Low\ntelescope model, we find constraints comparable to sample variance in the range\n$L<1000$, even for survey areas as small as $25\\mbox{ deg}^2$.\n",
"title": "Simulations for 21 cm radiation lensing at EoR redshifts"
}
| null | null | null | null | true | null |
8952
| null |
Default
| null | null |
null |
{
"abstract": " A self-repelling random walk of a token on a graph is one in which at each\nstep, the token moves to a neighbor that has been visited least often (with\nties broken randomly). The properties of self-repelling random walks have been\nanalyzed for two dimensional lattices and these walks have been shown to\nexhibit a remarkable uniformity with which they visit nodes in a graph. In this\npaper, we extend this analysis to self-repelling random walks on mobile\nnetworks in which the underlying graph itself is temporally evolving. Using\nnetwork simulations in ns-3, we characterize the number of times each node is\nvisited from the start until all nodes have been visited at least once. We\nevaluate under different mobility models and on networks ranging from 100 to\n1000 nodes. Our results show that until about 85% coverage, duplicate visits\nare very rare highlighting the efficiency with which a majority of nodes in the\nnetwork can be visited. Even at 100% coverage, the exploration overhead (the\nratio of number of steps to number of unique visited nodes) remains low and\nunder 2. Our analysis shows that self-repelling random walks are effective,\nstructure-free tools for data aggregation in mobile ad-hoc networks.\n",
"title": "Coverage characteristics of self-repelling random walks in mobile ad-hoc networks"
}
| null | null | null | null | true | null |
8953
| null |
Default
| null | null |
null |
{
"abstract": " We propose an algorithm to impute and forecast a time series by transforming\nthe observed time series into a matrix, utilizing matrix estimation to recover\nmissing values and de-noise observed entries, and performing linear regression\nto make predictions. At the core of our analysis is a representation result,\nwhich states that for a large model class, the transformed time series matrix\nis (approximately) low-rank. In effect, this generalizes the widely used\nSingular Spectrum Analysis (SSA) in time series literature, and allows us to\nestablish a rigorous link between time series analysis and matrix estimation.\nThe key to establishing this link is constructing a Page matrix with\nnon-overlapping entries rather than a Hankel matrix as is commonly done in the\nliterature (e.g., SSA). This particular matrix structure allows us to provide\nfinite sample analysis for imputation and prediction, and prove the asymptotic\nconsistency of our method. Another salient feature of our algorithm is that it\nis model agnostic with respect to both the underlying time dynamics and the\nnoise distribution in the observations. The noise agnostic property of our\napproach allows us to recover the latent states when only given access to noisy\nand partial observations a la a Hidden Markov Model; e.g., recovering the\ntime-varying parameter of a Poisson process without knowing that the underlying\nprocess is Poisson. Furthermore, since our forecasting algorithm requires\nregression with noisy features, our approach suggests a matrix estimation based\nmethod - coupled with a novel, non-standard matrix estimation error metric - to\nsolve the error-in-variable regression problem, which could be of interest in\nits own right. Through synthetic and real-world datasets, we demonstrate that\nour algorithm outperforms standard software packages (including R libraries) in\nthe presence of missing data as well as high levels of noise.\n",
"title": "Model Agnostic Time Series Analysis via Matrix Estimation"
}
| null | null | null | null | true | null |
8954
| null |
Default
| null | null |
null |
{
"abstract": " We consider the motion-planning problem of planning a collision-free path of\na robot in the presence of risk zones. The robot is allowed to travel in these\nzones but is penalized in a super-linear fashion for consecutive accumulative\ntime spent there. We suggest a natural cost function that balances path length\nand risk-exposure time. Specifically, we consider the discrete setting where we\nare given a graph, or a roadmap, and we wish to compute the minimal-cost path\nunder this cost function. Interestingly, paths defined using our cost function\ndo not have an optimal substructure. Namely, subpaths of an optimal path are\nnot necessarily optimal. Thus, the Bellman condition is not satisfied and\nstandard graph-search algorithms such as Dijkstra cannot be used. We present a\npath-finding algorithm, which can be seen as a natural generalization of\nDijkstra's algorithm. Our algorithm runs in $O\\left((n_B\\cdot n) \\log( n_B\\cdot\nn) + n_B\\cdot m\\right)$ time, where~$n$ and $m$ are the number of vertices and\nedges of the graph, respectively, and $n_B$ is the number of intersections\nbetween edges and the boundary of the risk zone. We present simulations on\nrobotic platforms demonstrating both the natural paths produced by our cost\nfunction and the computational efficiency of our algorithm.\n",
"title": "Efficient motion planning for problems lacking optimal substructure"
}
| null | null | null | null | true | null |
8955
| null |
Default
| null | null |
null |
{
"abstract": " Orthogonal frequency division multiplexing (OFDM) has been widely used in\ncommunication systems operating in the millimeter wave (mmWave) band to combat\nfrequency-selective fading and achieve multi-Gbps transmissions, such as IEEE\n802.15.3c and IEEE 802.11ad. For mmWave systems with ultra high sampling rate\nrequirements, the use of low-resolution analog-to-digital converters (ADCs)\n(i.e., 1-3 bits) ensures an acceptable level of power consumption and system\ncosts. However, orthogonality among sub-channels in the OFDM system cannot be\nmaintained because of the severe non-linearity caused by low-resolution ADC,\nwhich renders the design of data detector challenging. In this study, we\ndevelop an efficient algorithm for optimal data detection in the mmWave OFDM\nsystem with low-resolution ADCs. The analytical performance of the proposed\ndetector is derived and verified to achieve the fundamental limit of the\nBayesian optimal design. On the basis of the derived analytical expression, we\nfurther propose a power allocation (PA) scheme that seeks to minimize the\naverage symbol error rate. In addition to the optimal data detector, we also\ndevelop a feasible channel estimation method, which can provide high-quality\nchannel state information without significant pilot overhead. Simulation\nresults confirm the accuracy of our analysis and illustrate that the\nperformance of the proposed detector in conjunction with the proposed PA scheme\nis close to the optimal performance of the OFDM system with infinite-resolution\nADC.\n",
"title": "Bayesian Optimal Data Detector for mmWave OFDM System with Low-Resolution ADC"
}
| null | null | null | null | true | null |
8956
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study the generalized mean-field stochastic control problem\nwhen the usual stochastic maximum principle (SMP) is not applicable due to the\nsingularity of the Hamiltonian function. In this case, we derive a second order\nSMP. We introduce the adjoint process by the generalized mean-field backward\nstochastic differential equation. The keys in the proofs are the expansion of\nthe cost functional in terms of a perturbation parameter, and the use of the\nrange theorem for vector-valued measures.\n",
"title": "A second-order stochastic maximum principle for generalized mean-field control problem"
}
| null | null | null | null | true | null |
8957
| null |
Default
| null | null |
null |
{
"abstract": " We extend the Granger-Johansen representation theorems for I(1) and I(2)\nvector autoregressive processes to accommodate processes that take values in an\narbitrary complex separable Hilbert space. This more general setting is of\ncentral relevance for statistical applications involving functional time\nseries. We first obtain a range of necessary and sufficient conditions for a\npole in the inverse of a holomorphic index-zero Fredholm operator pencil to be\nof first or second order. Those conditions form the basis for our development\nof I(1) and I(2) representations of autoregressive Hilbertian processes.\nCointegrating and attractor subspaces are characterized in terms of the\nbehavior of the autoregressive operator pencil in a neighborhood of one.\n",
"title": "Representation of I(1) and I(2) autoregressive Hilbertian processes"
}
| null | null | null | null | true | null |
8958
| null |
Default
| null | null |
null |
{
"abstract": " The number density of field galaxies per rotation velocity, referred to as\nthe velocity function, is an intriguing statistical measure probing the\nsmallest scales of structure formation. In this paper we point out that the\nvelocity function is sensitive to small shifts in key cosmological parameters\nsuch as the amplitude of primordial perturbations ($\\sigma_8$) or the total\nmatter density ($\\Omega_{\\rm m}$). Using current data and applying conservative\nassumptions about baryonic effects, we show that the observed velocity function\nof the Local Volume favours cosmologies in tension with the measurements from\nPlanck but in agreement with the latest findings from weak lensing surveys.\nWhile the current systematics regarding the relation between observed and true\nrotation velocities are potentially important, upcoming data from HI surveys as\nwell as new insights from hydrodynamical simulations will dramatically improve\nthe situation in the near future.\n",
"title": "Constraining cosmology with the velocity function of low-mass galaxies"
}
| null | null | null | null | true | null |
8959
| null |
Default
| null | null |
null |
{
"abstract": " We study the signs of the Fourier coefficients of a newform. Let $f$ be a\nnormalized newform of weight $k$ for $\\Gamma_0(N)$. Let $a_f(n)$ be the $n$th\nFourier coefficient of $f$. For any fixed positive integer $m$, we study the\ndistribution of the signs of $\\{a_f(p^m)\\}_p$, where $p$ runs over all prime\nnumbers. We also find out the abscissas of absolute convergence of two\nDirichlet series with coefficients involving the Fourier coefficients of cusp\nforms and the coefficients of symmetric power $L$-functions.\n",
"title": "On the coefficients of symmetric power $L$-functions"
}
| null | null | null | null | true | null |
8960
| null |
Default
| null | null |
null |
{
"abstract": " Germanium telluride features special spin-electric effects originating from\nspin-orbit coupling and symmetry breaking by the ferroelectric lattice\npolarization, which opens up many prospectives for electrically tunable and\nswitchable spin electronic devices. By Mn doping of the {\\alpha}-GeTe host\nlattice, the system becomes a multiferroic semiconductor possessing\nmagnetoelectric properties in which the electric polarization, magnetization\nand spin texture are coupled to each other. Employing spin- and angle-resolved\nphotoemission spectroscopy in bulk- and surface-sensitive energy ranges and by\nvarying dipole transition matrix elements, we disentangle the bulk, surface and\nsurface-resonance states of the electronic structure and determine the spin\ntextures for selected parameters. From our results, we derive a comprehensive\nmodel of the {\\alpha}-GeTe surface electronic structure which fits experimental\ndata and first principle theoretical predictions and we discuss the\nunconventional evolution of the Rashba-type spin splitting upon manipulation by\nexternal B- and E-fields.\n",
"title": "Spin-resolved electronic structure of ferroelectric α-GeTe and multiferroic Ge1-xMnxTe"
}
| null | null | null | null | true | null |
8961
| null |
Default
| null | null |
null |
{
"abstract": " Objects may appear at arbitrary scales in perspective images of a scene,\nposing a challenge for recognition systems that process images at a fixed\nresolution. We propose a depth-aware gating module that adaptively selects the\npooling field size in a convolutional network architecture according to the\nobject scale (inversely proportional to the depth) so that small details are\npreserved for distant objects while larger receptive fields are used for those\nnearby. The depth gating signal is provided by stereo disparity or estimated\ndirectly from monocular input. We integrate this depth-aware gating into a\nrecurrent convolutional neural network to perform semantic segmentation. Our\nrecurrent module iteratively refines the segmentation results, leveraging the\ndepth and semantic predictions from the previous iterations.\nThrough extensive experiments on four popular large-scale RGB-D datasets, we\ndemonstrate this approach achieves competitive semantic segmentation\nperformance with a model which is substantially more compact. We carry out\nextensive analysis of this architecture including variants that operate on\nmonocular RGB but use depth as side-information during training, unsupervised\ngating as a generic attentional mechanism, and multi-resolution gating. We find\nthat gated pooling for joint semantic segmentation and depth yields\nstate-of-the-art results for quantitative monocular depth estimation.\n",
"title": "Recurrent Scene Parsing with Perspective Understanding in the Loop"
}
| null | null | null | null | true | null |
8962
| null |
Default
| null | null |
null |
{
"abstract": " The use of standard platforms in the field of humanoid robotics can\naccelerate research, and lower the entry barrier for new research groups. While\nmany affordable humanoid standard platforms exist in the lower size ranges of\nup to 60cm, beyond this the few available standard platforms quickly become\nsignificantly more expensive, and difficult to operate and maintain. In this\npaper, the igus Humanoid Open Platform is presented---a new, affordable,\nversatile and easily customisable standard platform for humanoid robots in the\nchild-sized range. At 90cm, the robot is large enough to interact with a\nhuman-scale environment in a meaningful way, and is equipped with enough torque\nand computing power to foster research in many possible directions. The\nstructure of the robot is entirely 3D printed, allowing for a lightweight and\nappealing design. The electrical and mechanical designs of the robot are\npresented, and the main features of the corresponding open-source ROS software\nare discussed. The 3D CAD files for all of the robot parts have been released\nopen-source in conjunction with this paper.\n",
"title": "Child-sized 3D Printed igus Humanoid Open Platform"
}
| null | null | null | null | true | null |
8963
| null |
Default
| null | null |
null |
{
"abstract": " This paper studies the complexity of solving two classes of non-cooperative\ngames in a distributed manner in which the players communicate with a set of\nsystem nodes over noisy communication channels. The complexity of solving each\ngame class is defined as the minimum number of iterations required to find a\nNash equilibrium (NE) of any game in that class with $\\epsilon$ accuracy.\nFirst, we consider the class $\\mathcal{G}$ of all $N$-player non-cooperative\ngames with a continuous action space that admit at least one NE. Using\ninformation-theoretic inequalities, we derive a lower bound on the complexity\nof solving $\\mathcal{G}$ that depends on the Kolmogorov $2\\epsilon$-capacity of\nthe constraint set and the total capacity of the communication channels. We\nalso derive a lower bound on the complexity of solving games in $\\mathcal{G}$\nwhich depends on the volume and surface area of the constraint set. We next\nconsider the class of all $N$-player non-cooperative games with at least one NE\nsuch that the players' utility functions satisfy a certain (differential)\nconstraint. We derive lower bounds on the complexity of solving this game class\nunder both Gaussian and non-Gaussian noise models. Our result in the\nnon-Gaussian case is derived by establishing a connection between the\nKullback-Leibler distance and Fisher information.\n",
"title": "Lower Bounds on the Complexity of Solving Two Classes of Non-cooperative Games"
}
| null | null | null | null | true | null |
8964
| null |
Default
| null | null |
null |
{
"abstract": " The angle between the spin of a star and its planets' orbital planes traces\nthe history of the planetary system. Exoplanets orbiting close to cool stars\nare expected to be on circular, aligned orbits because of strong tidal\ninteractions with the stellar convective envelope. Spin-orbit alignment can be\nmeasured when the planet transits its star, but such ground-based spectroscopic\nmeasurements are challenging for cool, slowly-rotating stars. Here we report\nthe characterization of a planet three-dimensional trajectory around an M dwarf\nstar, derived by mapping the spectrum of the stellar photosphere along the\nchord transited by the planet. We find that the eccentric orbit of the\nNeptune-mass exoplanet GJ 436b is nearly perpendicular to the stellar equator.\nBoth eccentricity and misalignment, surprising around a cool star, can result\nfrom dynamical interactions (via Kozai migration) with a yet-undetected outer\ncompanion. This inward migration of GJ 436b could have triggered the\natmospheric escape that now sustains its giant exosphere. Eccentric, misaligned\nexoplanets orbiting close to cool stars might thus hint at the presence of\nunseen perturbers and illustrate the diversity of orbital architectures seen in\nexoplanetary systems.\n",
"title": "Orbital misalignment of the Neptune-mass exoplanet GJ 436b with the spin of its cool star"
}
| null | null |
[
"Physics"
] | null | true | null |
8965
| null |
Validated
| null | null |
null |
{
"abstract": " We analysed the flux-flow region of isofield magneto resistivity data\nobtained on three crystals of BaFe$_{2-x}$Ni$_x$As$_2$ with $T_c$$\\sim$20 K for\nthree different geometries relative to the angle formed between the applied\nmagnetic field and the c-axis of the crystals. The field dependent activation\nenergy, $U_0$, was obtained from the TAFF and modified vortex-glass models,\nwhich were compared with the values of $U_0$ obtained from flux-creep available\nin the literature. We observed that the $U_0$ obtained from the TAFF model show\ndeviations among the different crystals, while the correspondent glass lines\nobtained from the vortex glass model are virtually coincident. It is shown that\nthe data is well explained by the modified vortex glass model, allowing to\nextract values of $T_g$, the glass transition temperature, and $T^*$, a\ntemperature which scales with the mean field critical temperature $T_c(H)$. The\nresulting glass lines obey the anisotropic Ginzburg-Landau theory and are well\nfitted by a theory developed in the literature by considering the effect of\ndisorder.\n",
"title": "Flux-flow and vortex-glass phase in iron pnictide BaFe$_{2-x}$Ni$_x$As$_2$ single crystals with $T_c$ $\\sim$ 20 K"
}
| null | null | null | null | true | null |
8966
| null |
Default
| null | null |
null |
{
"abstract": " Algorithmic issues concerning Elliott local semigroups are seldom considered\nin the literature, although these combinatorial structures completely classify\nAF algebras. In general, the addition operation of an Elliott local semigroup\nis {\\it partial}, but for every AF algebra $\\mathfrak B$ whose Murray-von\nNeumann order of projections is a lattice, this operation is uniquely\nextendible to the addition of an involutive monoid $E(\\mathfrak B)$. Let\n$\\mathfrak M_1$ be the Farey AF algebra introduced by the present author in\n1988 and rediscovered by F. Boca in 2008. The freeness properties of the\ninvolutive monoid $E(\\mathfrak M_1)$ yield a natural word problem for every AF\nalgebra $\\mathfrak B$ with singly generated $E(\\mathfrak B)$, because\n$\\mathfrak B$ is automatically a quotient of $\\mathfrak M_1$. Given two\nformulas $\\phi$ and $\\psi$ in the language of involutive monoids, the problem\nasks to decide whether $\\phi$ and $\\psi$ code the same equivalence of\nprojections of $\\mathfrak B$. This mimics the classical definition of the word\nproblem of a group presented by generators and relations. We show that the word\nproblem of $\\mathfrak M_1$ is solvable in polynomial time, and so is the word\nproblem of the Behnke-Leptin algebras $\\mathcal A_{n,k}$, and of the\nEffros-Shen algebras $\\mathfrak F_{\\theta}$, for $\\theta\\in [0,1]\\setminus\n\\mathbb Q$ a real algebraic number, or $\\theta = 1/e$. We construct a quotient\nof $\\mathfrak M_1$ having a Gödel incomplete word problem, and show that no\nprimitive quotient of $\\mathfrak M_1$ is Gödel incomplete.\n",
"title": "Word problems in Elliott monoids"
}
| null | null | null | null | true | null |
8967
| null |
Default
| null | null |
null |
{
"abstract": " We show that there is an absolute constant $c > 1/2$ such that the Mahler\nmeasure of the Fekete polynomials $f_p$ of the form $$f_p(z) :=\n\\sum_{k=1}^{p-1}{\\left( \\frac kp \\right)z^k}\\,,$$ (where the coefficients are\nthe usual Legendre symbols) is at least $c\\sqrt{p}$ for all sufficiently large\nprimes $p$. This improves the lower bound $\\left(\\frac 12 -\n\\varepsilon\\right)\\sqrt{p}$ known before for the Mahler measure of the Fekete\npolynomials $f_p$ for all sufficiently large primes $p \\geq c_{\\varepsilon}$.\nOur approach is based on the study of the zeros of the Fekete polynomials on\nthe unit circle.\n",
"title": "Improved lower bounds for the Mahler measure of the Fekete polynomials"
}
| null | null |
[
"Mathematics"
] | null | true | null |
8968
| null |
Validated
| null | null |
null |
{
"abstract": " Recently, supervised hashing methods have attracted much attention since they\ncan optimize retrieval speed and storage cost while preserving semantic\ninformation. Because hashing codes learning is NP-hard, many methods resort to\nsome form of relaxation technique. But the performance of these methods can\neasily deteriorate due to the relaxation. Luckily, many supervised hashing\nformulations can be viewed as energy functions, hence solving hashing codes is\nequivalent to learning marginals in the corresponding conditional random field\n(CRF). By minimizing the KL divergence between a fully factorized distribution\nand the Gibbs distribution of this CRF, a set of consistency equations can be\nobtained, but updating them in parallel may not yield a local optimum since the\nvariational lower bound is not guaranteed to increase. In this paper, we use a\nlinear approximation of the sigmoid function to convert these consistency\nequations to linear systems, which have a closed-form solution. By applying\nthis novel technique to two classical hashing formulations KSH and SPLH, we\nobtain two new methods called EM (energy minimizing based)-KSH and EM-SPLH.\nExperimental results on three datasets show the superiority of our methods.\n",
"title": "Supervised Hashing based on Energy Minimization"
}
| null | null | null | null | true | null |
8969
| null |
Default
| null | null |
null |
{
"abstract": " The paper provides results for the application of boundary feedback control\nwith Zero-Order-Hold (ZOH) to 1-D linear, first-order, hyperbolic systems with\nnon-local terms on bounded domains. It is shown that the emulation design based\non the recently proposed continuous-time, boundary feedback, designed by means\nof backstepping, guarantees closed-loop exponential stability, provided that\nthe sampling period is sufficiently small. It is also shown that, contrary to\nthe parabolic case, a smaller sampling period implies a faster convergence rate\nwith no upper bound for the achieved convergence rate. The obtained results\nprovide stability estimates for the sup-norm of the state and robustness with\nrespect to perturbations of the sampling schedule is guaranteed.\n",
"title": "Sampled-Data Boundary Feedback Control of 1-D Hyperbolic PDEs with Non-Local Terms"
}
| null | null | null | null | true | null |
8970
| null |
Default
| null | null |
null |
{
"abstract": " Accurate and efficient entity resolution is an open challenge of particular\nrelevance to intelligence organisations that collect large datasets from\ndisparate sources with differing levels of quality and standard. Starting from\na first-principles formulation of entity resolution, this paper presents a\nnovel Entity Resolution algorithm that introduces a data-driven blocking and\nrecord-linkage technique based on the probabilistic identification of entity\nsignatures in data. The scalability and accuracy of the proposed algorithm are\nevaluated using benchmark datasets and shown to achieve state-of-the-art\nresults. The proposed algorithm can be implemented simply on modern parallel\ndatabases, which allows it to be deployed with relative ease in large\nindustrial applications.\n",
"title": "Scalable Entity Resolution Using Probabilistic Signatures on Parallel Databases"
}
| null | null | null | null | true | null |
8971
| null |
Default
| null | null |
null |
{
"abstract": " The admittance of two types of Josephson weak links is calculated, i.e., of a\none-dimensional superconducting wire with a local suppression of the order\nparameter, and the second is a short S-c-S structure, where S denotes a\nsuperconductor and c---a constriction. The systems of the first type are\nanalyzed on the basis of time-dependent Ginzburg-Landau equations. We show that\nthe impedance $Z(\\Omega)$ has a maximum as a function of the frequency\n$\\Omega$, and the electric field $E_{\\Omega}$ is determined by two\ngauge-invariant quantities---the condensate momentum $Q_{\\Omega}$ and the\npotential $\\mu$ related to charge imbalance. The structures of the second type\nare studied on the basis of microscopic equations for quasiclassical Green's\nfunctions in the Keldysh technique. For short S-c-S contacts (the Thouless\nenergy ${E_{\\text{Th}} = D/L^{2} \\gg \\Delta}$) we present a formula for\nadmittance $Y$ valid at frequencies $\\Omega$ and temperatures $T$ less than the\nThouless energy but arbitrary with respect to the energy gap $\\Delta$. It is\nshown that, at low temperatures, the absorption is absent [${\\mathrm{Re}(Y) =\n0}$] if the frequency does not exceed the energy gap in the center of the\nconstriction (${\\Omega < \\Delta \\cos \\varphi_{0}}$, where $2 \\varphi_{0}$ is\nthe phase difference between the S reservoirs). The absorption gradually\nincreases with increasing the difference ${(\\Omega - \\Delta \\cos \\varphi_{0})}$\nif $2 \\varphi_{0}$ is less than the phase difference $2 \\varphi_{\\text{c}}$\ncorresponding to the critical Josephson current. In the interval ${2\n\\varphi_{\\text{c}} < 2 \\varphi_{0} < \\pi}$, the absorption has a maximum. This\ninterval of the phase difference is achievable in phase-biased Josephson\njunctions. Close to $T_{\\text{c}}$ the admittance has a maximum at low $\\Omega$\nwhich is described by an analytical formula.\n",
"title": "ac properties of short Josephson weak links"
}
| null | null |
[
"Physics"
] | null | true | null |
8972
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper we propose an efficient algorithm ProtoDash for selecting\nprototypical examples from complex datasets. Our generalizes the learn to\ncriticize (L2C) work by Kim et al. (2016) to not only select prototypes for a\ngiven sparsity level $m$ but also to associate non-negative (for\ninterpretability) weights with each of them indicative of the importance of\neach prototype. This extension provides a single coherent framework under which\nboth prototypes and criticisms can be found. Furthermore, our framework works\nfor any symmetric positive definite kernel thus addressing one of the key open\nquestions laid out in Kim et al. (2016). Our additional requirement of learning\nnon-negative weights no longer maintains submodularity of the objective as in\nthe previous work, however, we show that the problem is weakly submodular and\nderive approximation guarantees for our fast ProtoDash algorithm. We\ndemonstrate the efficacy of our method on diverse domains such as retail, digit\nrecognition (MNIST) and on publicly available 40 health questionnaires obtained\nfrom the Center for Disease Control (CDC) website maintained by the US Dept. of\nHealth. We validate the results quantitatively as well as qualitatively based\non expert feedback and recently published scientific studies on public health,\nthus showcasing the power of our method in providing actionability (for\nretail), utility (for MNIST) and insight (on CDC datasets), which presumably\nare the hallmark of an effective interpretable method.\n",
"title": "ProtoDash: Fast Interpretable Prototype Selection"
}
| null | null | null | null | true | null |
8973
| null |
Default
| null | null |
null |
{
"abstract": " We consider a variant of the classic multi-armed bandit problem where the\nexpected reward of each arm is a function of an unknown parameter. The arms are\ndivided into different groups, each of which has a common parameter. Therefore,\nwhen the player selects an arm at each time slot, information of other arms in\nthe same group is also revealed. This regional bandit model naturally bridges\nthe non-informative bandit setting where the player can only learn the chosen\narm, and the global bandit model where sampling one arms reveals information of\nall arms. We propose an efficient algorithm, UCB-g, that solves the regional\nbandit problem by combining the Upper Confidence Bound (UCB) and greedy\nprinciples. Both parameter-dependent and parameter-free regret upper bounds are\nderived. We also establish a matching lower bound, which proves the\norder-optimality of UCB-g. Moreover, we propose SW-UCB-g, which is an extension\nof UCB-g for a non-stationary environment where the parameters slowly vary over\ntime.\n",
"title": "Regional Multi-Armed Bandits"
}
| null | null | null | null | true | null |
8974
| null |
Default
| null | null |
null |
{
"abstract": " Asteroseismic parameters allow us to measure the basic stellar properties of\nfield giants observed far across the Galaxy. Most of such determinations are,\nup to now, based on simple scaling relations involving the large frequency\nseparation, \\Delta\\nu, and the frequency of maximum power, \\nu$_{max}$. In this\nwork, we implement \\Delta\\nu\\ and the period spacing, {\\Delta}P, computed along\ndetailed grids of stellar evolutionary tracks, into stellar isochrones and\nhence in a Bayesian method of parameter estimation. Tests with synthetic data\nreveal that masses and ages can be determined with typical precision of 5 and\n19 per cent, respectively, provided precise seismic parameters are available.\nAdding independent information on the stellar luminosity, these values can\ndecrease down to 3 and 10 per cent respectively. The application of these\nmethods to NGC 6819 giants produces a mean age in agreement with those derived\nfrom isochrone fitting, and no evidence of systematic differences between RGB\nand RC stars. The age dispersion of NGC 6819 stars, however, is larger than\nexpected, with at least part of the spread ascribable to stars that underwent\nmass-transfer events.\n",
"title": "Determining stellar parameters of asteroseismic targets: going beyond the use of scaling relations"
}
| null | null | null | null | true | null |
8975
| null |
Default
| null | null |
null |
{
"abstract": " In the polarised Drell-Yan experiment at the COMPASS facility in CERN pion\nbeam with momentum of 190 GeV/c and intensity about $10^8$ pions/s interacted\nwith transversely polarised NH$_3$ target. Muon pairs produced in Drel-Yan\nprocess were detected. The measurement was done in 2015 as the 1st ever\npolarised Drell-Yan fixed target experiment. The hydrogen nuclei in the\nsolid-state NH$_3$ were polarised by dynamic nuclear polarisation in 2.5 T\nfield of large-acceptance superconducting magnet. Large helium dilution\ncryostat was used to cool the target down below 100 mK. Polarisation of\nhydrogen nuclei reached during the data taking was about 80 %. Two oppositely\npolarised target cells, each 55 cm long and 4 cm in diameter were used.\nOverview of COMPASS facility and the polarised target with emphasis on the\ndilution cryostat and magnet is given. Results of the polarisation measurement\nin the Drell-Yan run and overviews of the target material, cell and dynamic\nnuclear polarisation system are given in the part II.\n",
"title": "Polarised target for Drell-Yan experiment in COMPASS at CERN, part I"
}
| null | null | null | null | true | null |
8976
| null |
Default
| null | null |
null |
{
"abstract": " Early approaches to multiple-output Gaussian processes (MOGPs) relied on\nlinear combinations of independent, latent, single-output Gaussian processes\n(GPs). This resulted in cross-covariance functions with limited parametric\ninterpretation, thus conflicting with the ability of single-output GPs to\nunderstand lengthscales, frequencies and magnitudes to name a few. On the\ncontrary, current approaches to MOGP are able to better interpret the\nrelationship between different channels by directly modelling the\ncross-covariances as a spectral mixture kernel with a phase shift. We extend\nthis rationale and propose a parametric family of complex-valued cross-spectral\ndensities and then build on Cramér's Theorem (the multivariate version of\nBochner's Theorem) to provide a principled approach to design multivariate\ncovariance functions. The so-constructed kernels are able to model delays among\nchannels in addition to phase differences and are thus more expressive than\nprevious methods, while also providing full parametric interpretation of the\nrelationship across channels. The proposed method is first validated on\nsynthetic data and then compared to existing MOGP methods on two real-world\nexamples.\n",
"title": "Spectral Mixture Kernels for Multi-Output Gaussian Processes"
}
| null | null | null | null | true | null |
8977
| null |
Default
| null | null |
null |
{
"abstract": " A personal recollection of events that preceded the construction of\nSupergravity and of some subsequent developments.\n",
"title": "Supergravity and its Legacy"
}
| null | null | null | null | true | null |
8978
| null |
Default
| null | null |
null |
{
"abstract": " We demonstrate that the five vortex equations recently introduced by Manton\nariseas symmetry reductions of the anti-self-dual Yang--Mills equations in four\ndimensions. In particular the Jackiw--Pi vortex and the Ambj\\o rn--Olesen\nvortex correspond to the gauge group $SU(1, 1)$, and respectively the Euclidean\nor the $SU(2)$ symmetry groups acting with two-dimensional orbits. We show how\nto obtain vortices with higher vortex numbers, by superposing vortex equations\nof different types. Finally we use the kinetic energy of the Yang--Mills theory\nin 4+1 dimensions to construct a metric on vortex moduli spaces. This metric is\nnot positive-definite in cases of non-compact gauge groups.\n",
"title": "Manton's five vortex equations from self-duality"
}
| null | null | null | null | true | null |
8979
| null |
Default
| null | null |
null |
{
"abstract": " Humans develop a common sense of style compatibility between items based on\ntheir attributes. We seek to automatically answer questions like \"Does this\nshirt go well with that pair of jeans?\" In order to answer these kinds of\nquestions, we attempt to model human sense of style compatibility in this\npaper. The basic assumption of our approach is that most of the important\nattributes for a product in an online store are included in its title\ndescription. Therefore it is feasible to learn style compatibility from these\ndescriptions. We design a Siamese Convolutional Neural Network architecture and\nfeed it with title pairs of items, which are either compatible or incompatible.\nThose pairs will be mapped from the original space of symbolic words into some\nembedded style space. Our approach takes only words as the input with few\npreprocessing and there is no laborious and expensive feature engineering.\n",
"title": "Deep Style Match for Complementary Recommendation"
}
| null | null | null | null | true | null |
8980
| null |
Default
| null | null |
null |
{
"abstract": " We study the problem of learning a latent variable model from a stream of\ndata. Latent variable models are popular in practice because they can explain\nobserved data in terms of unobserved concepts. These models have been\ntraditionally studied in the offline setting. In the online setting, on the\nother hand, the online EM is arguably the most popular algorithm for learning\nlatent variable models. Although the online EM is computationally efficient, it\ntypically converges to a local optimum. In this work, we develop a new online\nlearning algorithm for latent variable models, which we call SpectralLeader.\nSpectralLeader always converges to the global optimum, and we derive a\nsublinear upper bound on its $n$-step regret in the bag-of-words model. In both\nsynthetic and real-world experiments, we show that SpectralLeader performs\nsimilarly to or better than the online EM with tuned hyper-parameters.\n",
"title": "SpectralLeader: Online Spectral Learning for Single Topic Models"
}
| null | null | null | null | true | null |
8981
| null |
Default
| null | null |
null |
{
"abstract": " We aim to introduce the generalized multiindex Bessel function $J_{\\left(\n\\beta _{j}\\right) _{m},\\kappa ,b}^{\\left( \\alpha _{j}\\right)_{m},\\gamma\n,c}\\left[ z\\right] $ and to present some formulas of the Riemann-Liouville\nfractional integration and differentiation operators. Further, we also derive\ncertain integral formulas involving the newly defined generalized multiindex\nBessel function $J_{\\left( \\beta _{j}\\right) _{m},\\kappa ,b}^{\\left( \\alpha\n_{j}\\right)_{m},\\gamma ,c}\\left[ z\\right] $. We prove that such integrals are\nexpressed in terms of the Fox-Wright function $_{p}\\Psi_{q}(z)$. The results\npresented here are of general in nature and easily reducible to new and known\nresults.\n",
"title": "Fractional Calculus and certain integrals of Generalized multiindex Bessel function"
}
| null | null | null | null | true | null |
8982
| null |
Default
| null | null |
null |
{
"abstract": " We show, assuming a mild set-theoretic hypothesis, that if an abstract\nelementary class (AEC) has a superstable-like forking notion for models of\ncardinality $\\lambda$ and a superstable-like forking notion for models of\ncardinality $\\lambda^+$, then orbital types over models of cardinality\n$\\lambda^+$ are determined by their restrictions to submodels of cardinality\n$\\lambda$. By a superstable-like forking notion, we mean here a good frame, a\ncentral concept of Shelah's book on AECs.\nIt is known that locality of orbital types together with the existence of a\nsuperstable-like notion for models of cardinality $\\lambda$ implies the\nexistence of a superstable-like notion for models of cardinality $\\lambda^+$,\nbut here we prove the converse. An immediate consequence is that forking in\n$\\lambda^+$ can be described in terms of forking in $\\lambda$.\n",
"title": "Tameness from two successive good frames"
}
| null | null | null | null | true | null |
8983
| null |
Default
| null | null |
null |
{
"abstract": " Motivated by the ${\\rm \\Psi}$-Riemann-Liouville $({\\rm \\Psi-RL})$ fractional\nderivative and by the ${\\rm \\Psi}$-Hilfer $({\\rm \\Psi-H})$ fractional\nderivative, we introduced a new fractional operator the so-called\n$\\rm\\Psi-$fractional integral. We present some important results by means of\ntheorems and in particular, that the $\\rm\\Psi-$fractional integration operator\nis limited. In this sense, we discuss some examples, in particular, involving\nthe Mittag-Leffler $({\\rm M-L})$ function, of paramount importance in the\nsolution of population growth problem, as approached. On the other hand, we\nrealize a brief discussion on the uniqueness of nonlinear $\\Psi$-fractional\nVolterra integral equation (${\\rm VIE}$) using $\\beta-$distance functions.\n",
"title": "On the $\\mathbf{\\rmΨ}-$fractional integral and applications"
}
| null | null | null | null | true | null |
8984
| null |
Default
| null | null |
null |
{
"abstract": " Illicit online pharmacies allow the purchase of prescription drugs online\nwithout a prescription. Such pharmacies leverage social media platforms such as\nTwit- ter as a promotion and marketing tool with the intent of reaching out to\na larger, potentially younger demographics of the population. Given the serious\nnegative health effects that arise from abusing such drugs, it is important to\nidentify the relevant content on social media and exterminate their presence as\nquickly as pos- sible. In response, we collected all the tweets that contained\nthe names of certain preselected controlled substances over a period of 5\nmonths. We found that an unsupervised topic modeling based methodology is able\nto identify tweets that promote and market controlled substances with high\nprecision. We also study the meta-data characteristics of such tweets and the\nusers who post them and find that they have several distinguishing\ncharacteristics that sets them apart. We were able to train supervised methods\nand achieve high performance in detecting such content and the users who post\nthem.\n",
"title": "Detection and Characterization of Illegal Marketing and Promotion of Prescription Drugs on Twitter"
}
| null | null | null | null | true | null |
8985
| null |
Default
| null | null |
null |
{
"abstract": " We use the dimension and the Lie algebra structure of the first Hochschild\ncohomology group to distinguish some algebras of dihedral, semi-dihedral and\nquaternion type up to stable equivalence of Morita type. In particular, we\ncomplete the classification of algebras of dihedral type that was mostly\ndetermined by Zhou and Zimmermann.\n",
"title": "First Hochschild cohomology group and stable equivalence classification of Morita type of some tame symmetric algebras"
}
| null | null |
[
"Mathematics"
] | null | true | null |
8986
| null |
Validated
| null | null |
null |
{
"abstract": " We analyze the convergence of (stochastic) gradient descent algorithm for\nlearning a convolutional filter with Rectified Linear Unit (ReLU) activation\nfunction. Our analysis does not rely on any specific form of the input\ndistribution and our proofs only use the definition of ReLU, in contrast with\nprevious works that are restricted to standard Gaussian input. We show that\n(stochastic) gradient descent with random initialization can learn the\nconvolutional filter in polynomial time and the convergence rate depends on the\nsmoothness of the input distribution and the closeness of patches. To the best\nof our knowledge, this is the first recovery guarantee of gradient-based\nalgorithms for convolutional filter on non-Gaussian input distributions. Our\ntheory also justifies the two-stage learning rate strategy in deep neural\nnetworks. While our focus is theoretical, we also present experiments that\nillustrate our theoretical findings.\n",
"title": "When is a Convolutional Filter Easy To Learn?"
}
| null | null | null | null | true | null |
8987
| null |
Default
| null | null |
null |
{
"abstract": " A large body of compelling evidence has been accumulated demonstrating that\nembodiment - the agent's physical setup, including its shape, materials,\nsensors and actuators - is constitutive for any form of cognition and as a\nconsequence, models of cognition need to be embodied. In contrast to methods\nfrom empirical sciences to study cognition, robots can be freely manipulated\nand virtually all key variables of their embodiment and control programs can be\nsystematically varied. As such, they provide an extremely powerful tool of\ninvestigation. We present a robotic bottom-up or developmental approach,\nfocusing on three stages: (a) low-level behaviors like walking and reflexes,\n(b) learning regularities in sensorimotor spaces, and (c) human-like cognition.\nWe also show that robotic based research is not only a productive path to\ndeepening our understanding of cognition, but that robots can strongly benefit\nfrom human-like cognition in order to become more autonomous, robust,\nresilient, and safe.\n",
"title": "Robots as Powerful Allies for the Study of Embodied Cognition from the Bottom Up"
}
| null | null | null | null | true | null |
8988
| null |
Default
| null | null |
null |
{
"abstract": " Skyrmions are topologically protected, two-dimensional, localized hedgehogs\nand whorls of spin. Originally invented as a concept in field theory for\nnuclear interactions, skyrmions are central to a wide range of phenomena in\ncondensed matter. Their realization at room temperature (RT) in magnetic\nmultilayers has generated considerable interest, fueled by technological\nprospects and the access granted to fundamental questions. The interaction of\nskyrmions with charge carriers gives rise to exotic electrodynamics, such as\nthe topological Hall effect (THE), the Hall response to an emergent magnetic\nfield, a manifestation of the skyrmion Berry-phase. The proposal that THE can\nbe used to detect skyrmions needs to be tested quantitatively. For that it is\nimperative to develop comprehensive understanding of skyrmions and other chiral\ntextures, and their electrical fingerprint. Here, using Hall transport and\nmagnetic imaging, we track the evolution of magnetic textures and their THE\nsignature in a technologically viable multilayer film as a function of\ntemperature ($T$) and out-of-plane applied magnetic field ($H$). We show that\ntopological Hall resistivity ($\\rho_\\mathrm{TH}$) scales with the density of\nisolated skyrmions ($n_\\mathrm{sk}$) over a wide range of $T$, confirming the\nimpact of the skyrmion Berry-phase on electronic transport. We find that at\nhigher $n_\\mathrm{sk}$ skyrmions cluster into worms which carry considerable\ntopological charge, unlike topologically-trivial spin spirals. While we\nestablish a qualitative agreement between $\\rho_\\mathrm{TH}(H,T)$ and areal\ndensity of topological charge $n_\\mathrm{T}(H,T)$, our detailed quantitative\nanalysis shows a much larger $\\rho_\\mathrm{TH}$ than the prevailing theory\npredicts for observed $n_\\mathrm{T}$.\n",
"title": "Chiral magnetic textures in Ir/Fe/Co/Pt multilayers: Evolution and topological Hall signature"
}
| null | null | null | null | true | null |
8989
| null |
Default
| null | null |
null |
{
"abstract": " Using a modification of the Shapiro scaling approach, we derive the\ndistribution of conductance in the magnetic field applicable in the vicinity of\nthe Anderson transition. This distribution is described by the same equations\nas in the absence of a field. Variation of the magnetic field does not lead to\nany qualitative effects in the conductance distribution and only changes its\nquantitative characteristics, moving a position of the system in the\nthree-parameter space. In contrast to the original Shapiro approach, the\nevolution equation for quasi-1D systems is established from the generalized\nDMPK equation, and not by a simple analogy with one-dimensional systems; as a\nresult, the whole approach became more rigorous and accurate.\n",
"title": "Conductance distribution in the magnetic field"
}
| null | null | null | null | true | null |
8990
| null |
Default
| null | null |
null |
{
"abstract": " Phase retrieval has been an attractive but difficult problem rising from\nphysical science, and there has been a gap between state-of-the-art theoretical\nconvergence analyses and the corresponding efficient retrieval methods.\nFirstly, these analyses all assume that the sensing vectors and the iterative\nupdates are independent, which only fits the ideal model with infinite\nmeasurements but not the reality, where data are limited and have to be reused.\nSecondly, the empirical results of some efficient methods, such as the\nrandomized Kaczmarz method, show linear convergence, which is beyond existing\ntheoretical explanations considering its randomness and reuse of data. In this\nwork, we study for the first time, without the independence assumption, the\nconvergence behavior of the randomized Kaczmarz method for phase retrieval.\nSpecifically, beginning from taking expectation of the squared estimation error\nwith respect to the index of measurement by fixing the sensing vector and the\nerror in the previous step, we discard the independence assumption, rigorously\nderive the upper and lower bounds of the reduction of the mean squared error,\nand prove the linear convergence. This work fills the gap between a fast\nconverging algorithm and its theoretical understanding. The proposed\nmethodology may contribute to the study of other iterative algorithms for phase\nretrieval and other problems in the broad area of signal processing and machine\nlearning.\n",
"title": "Linear Convergence of An Iterative Phase Retrieval Algorithm with Data Reuse"
}
| null | null | null | null | true | null |
8991
| null |
Default
| null | null |
null |
{
"abstract": " We show that any smooth bi-Lipschitz $h$ can be represented exactly as a\ncomposition $h_m \\circ ... \\circ h_1$ of functions $h_1,...,h_m$ that are close\nto the identity in the sense that each $\\left(h_i-\\mathrm{Id}\\right)$ is\nLipschitz, and the Lipschitz constant decreases inversely with the number $m$\nof functions composed. This implies that $h$ can be represented to any accuracy\nby a deep residual network whose nonlinear layers compute functions with a\nsmall Lipschitz constant. Next, we consider nonlinear regression with a\ncomposition of near-identity nonlinear maps. We show that, regarding Fréchet\nderivatives with respect to the $h_1,...,h_m$, any critical point of a\nquadratic criterion in this near-identity region must be a global minimizer. In\ncontrast, if we consider derivatives with respect to parameters of a fixed-size\nresidual network with sigmoid activation functions, we show that there are\nnear-identity critical points that are suboptimal, even in the realizable case.\nInformally, this means that functional gradient methods for residual networks\ncannot get stuck at suboptimal critical points corresponding to near-identity\nlayers, whereas parametric gradient methods for sigmoidal residual networks\nsuffer from suboptimal critical points in the near-identity region.\n",
"title": "Representing smooth functions as compositions of near-identity functions with implications for deep network optimization"
}
| null | null | null | null | true | null |
8992
| null |
Default
| null | null |
null |
{
"abstract": " We start with the recently conjectured 3d bosonization dualities and gauge\nglobal symmetries to generate an infinite sequence of new dualities. These\nequate theories with non-Abelian product gauge groups and bifundamental matter.\nWe uncover examples of Bose/Bose and Fermi/Fermi dualities, as well as a\nsequence of dualities between theories with scalar matter in two-index\nrepresentations. Our conjectures are consistent with level/rank duality in\nmassive phases.\n",
"title": "Bosonizing three-dimensional quiver gauge theories"
}
| null | null |
[
"Physics"
] | null | true | null |
8993
| null |
Validated
| null | null |
null |
{
"abstract": " Deep learning has emerged as a powerful machine learning technique that\nlearns multiple layers of representations or features of the data and produces\nstate-of-the-art prediction results. Along with the success of deep learning in\nmany other application domains, deep learning is also popularly used in\nsentiment analysis in recent years. This paper first gives an overview of deep\nlearning and then provides a comprehensive survey of its current applications\nin sentiment analysis.\n",
"title": "Deep Learning for Sentiment Analysis : A Survey"
}
| null | null | null | null | true | null |
8994
| null |
Default
| null | null |
null |
{
"abstract": " Differential privacy mechanisms that also make reconstruction of the data\nimpossible come at a cost - a decrease in utility. In this paper, we tackle\nthis problem by designing a private data release mechanism that makes\nreconstruction of the original data impossible and also preserves utility for a\nwide range of machine learning algorithms. We do so by combining the\nJohnson-Lindenstrauss (JL) transform with noise generated from a Laplace\ndistribution. While the JL transform can itself provide privacy guarantees\n\\cite{blocki2012johnson} and make reconstruction impossible, we do not rely on\nits differential privacy properties and only utilize its ability to make\nreconstruction impossible. We present novel proofs to show that our mechanism\nis differentially private under single element changes as well as single row\nchanges to any database. In order to show utility, we prove that our mechanism\nmaintains pairwise distances between points in expectation and also show that\nits variance is proportional to the the dimensionality of the subspace we\nproject the data into. Finally, we experimentally show the utility of our\nmechanism by deploying it on the task of clustering.\n",
"title": "Utility Preserving Secure Private Data Release"
}
| null | null | null | null | true | null |
8995
| null |
Default
| null | null |
null |
{
"abstract": " Employing the spin degree of freedom of charge carriers offers the\npossibility to extend the functionality of conventional electronic devices,\nwhile colloidal chemistry can be used to synthesize inexpensive and tuneable\nnanomaterials. In order to benefit from both concepts, Rashba spin-orbit\ninteraction has been investigated in colloidal lead sulphide nanosheets by\nelectrical measurements on the circular photo-galvanic effect. Lead sulphide\nnanosheets possess rock salt crystal structure, which is centrosymmetric. The\nsymmetry can be broken by quantum confinement, asymmetric vertical interfaces\nand a gate electric field leading to Rashba-type band splitting in momentum\nspace at the M points, which results in an unconventional selection mechanism\nfor the excitation of the carriers. The effect, which is supported by\nsimulations of the band structure using density functional theory, can be tuned\nby the gate electric field and by the thickness of the sheets. Spin-related\nelectrical transport phenomena in colloidal materials open a promising pathway\ntowards future inexpensive spintronic devices.\n",
"title": "Towards colloidal spintronics through Rashba spin-orbit interaction in lead sulphide nanosheets"
}
| null | null | null | null | true | null |
8996
| null |
Default
| null | null |
null |
{
"abstract": " In biodiversity and ecosystem functioning (BEF) research, the Loreau-Hector\n(LH) statistical scheme is widely-used to partition the effect of biodiversity\non ecosystem properties into a \"complementarity effect\" and a \"selection\neffect\". This selection effect was originally considered analogous to the\nselection term in the Price equation from evolutionary biology. However, a key\npaper published over thirteen years ago challenged this interpretation by\ndevising a new tripartite partitioning scheme that purportedly quantified the\nrole of selection in biodiversity experiments more accurately. This tripartite\nmethod, as well as its recent spatiotemporal extension, were both developed as\nan attempt to apply the Price equation in a BEF context. Here, we demonstrate\nthat the derivation of this tripartite method, as well as its spatiotemporal\nextension, involve a set of incoherent and nonsensical mathematical arguments\ndriven largely by naïve visual analogies with the original Price equation,\nthat result in neither partitioning scheme quantifying any real property in the\nnatural world. Furthermore, we show that Loreau and Hector's original selection\neffect always represented a true analog of the original Price selection term,\nmaking the tripartite partitioning scheme a nonsensical solution to a\nnon-existent problem [...]\n",
"title": "On the use and abuse of Price equation concepts in ecology"
}
| null | null | null | null | true | null |
8997
| null |
Default
| null | null |
null |
{
"abstract": " We have developed a data-driven magnetohydrodynamic (MHD) model of the global\nsolar corona which uses characteristically-consistent boundary conditions (BCs)\nat the inner boundary. Our global solar corona model can be driven by different\nobservational data including Solar Dynamics Observatory/Helioseismic and\nMagnetic Imager (SDO/HMI) synoptic vector magnetograms together with the\nhorizontal velocity data in the photosphere obtained by the time-distance\nhelioseismology method, and the line-of-sight (LOS) magnetogram data obtained\nby HMI, Solar and Heliospheric Observatory/Michelson Doppler Imager (SOHO/MDI),\nNational Solar Observatory/Global Oscillation Network Group (NSO/GONG) and\nWilcox Solar Observatory (WSO). We implemented our model in the Multi-Scale\nFluid-Kinetic Simulation Suite (MS-FLUKSS) - a suite of adaptive mesh\nrefinement (AMR) codes built upon the Chombo AMR framework developed at the\nLawrence Berkeley National Laboratory. We present an overview of our model,\ncharacteristic BCs, and two results we obtained using our model: A benchmark\ntest of relaxation of a dipole field using characteristic BCs, and relaxation\nof an initial PFSS field driven by HMI LOS magnetogram data, and horizontal\nvelocity data obtained by the time-distance helioseismology method using a set\nof non-characteristic BCs.\n",
"title": "A Data-Driven MHD Model of the Global Solar Corona within Multi-Scale Fluid-Kinetic Simulation Suite (MS-FLUKSS)"
}
| null | null | null | null | true | null |
8998
| null |
Default
| null | null |
null |
{
"abstract": " Neural network based approximate computing is a universal architecture\npromising to gain tremendous energy-efficiency for many error resilient\napplications. To guarantee the approximation quality, existing works deploy two\nneural networks (NNs), e.g., an approximator and a predictor. The approximator\nprovides the approximate results, while the predictor predicts whether the\ninput data is safe to approximate with the given quality requirement. However,\nit is non-trivial and time-consuming to make these two neural network\ncoordinate---they have different optimization objectives---by training them\nseparately. This paper proposes a novel neural network structure---AXNet---to\nfuse two NNs to a holistic end-to-end trainable NN. Leveraging the philosophy\nof multi-task learning, AXNet can tremendously improve the invocation\n(proportion of safe-to-approximate samples) and reduce the approximation error.\nThe training effort also decrease significantly. Experiment results show 50.7%\nmore invocation and substantial cuts of training time when compared to existing\nneural network based approximate computing framework.\n",
"title": "AXNet: ApproXimate computing using an end-to-end trainable neural network"
}
| null | null |
[
"Statistics"
] | null | true | null |
8999
| null |
Validated
| null | null |
null |
{
"abstract": " Deep learning methods are useful for high-dimensional data and are becoming\nwidely used in many areas of software engineering. Deep learners utilizes\nextensive computational power and can take a long time to train-- making it\ndifficult to widely validate and repeat and improve their results. Further,\nthey are not the best solution in all domains. For example, recent results show\nthat for finding related Stack Overflow posts, a tuned SVM performs similarly\nto a deep learner, but is significantly faster to train. This paper extends\nthat recent result by clustering the dataset, then tuning very learners within\neach cluster. This approach is over 500 times faster than deep learning (and\nover 900 times faster if we use all the cores on a standard laptop computer).\nSignificantly, this faster approach generates classifiers nearly as good\n(within 2\\% F1 Score) as the much slower deep learning method. Hence we\nrecommend this faster methods since it is much easier to reproduce and utilizes\nfar fewer CPU resources. More generally, we recommend that before researchers\nrelease research results, that they compare their supposedly sophisticated\nmethods against simpler alternatives (e.g applying simpler learners to build\nlocal models).\n",
"title": "500+ Times Faster Than Deep Learning (A Case Study Exploring Faster Methods for Text Mining StackOverflow)"
}
| null | null | null | null | true | null |
9000
| null |
Default
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.