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": " Neural speech synthesis models have recently demonstrated the ability to\nsynthesize high quality speech for text-to-speech and compression applications.\nThese new models often require powerful GPUs to achieve real-time operation, so\nbeing able to reduce their complexity would open the way for many new\napplications. We propose LPCNet, a WaveRNN variant that combines linear\nprediction with recurrent neural networks to significantly improve the\nefficiency of speech synthesis. We demonstrate that LPCNet can achieve\nsignificantly higher quality than WaveRNN for the same network size and that\nhigh quality LPCNet speech synthesis is achievable with a complexity under 3\nGFLOPS. This makes it easier to deploy neural synthesis applications on\nlower-power devices, such as embedded systems and mobile phones.\n",
"title": "LPCNet: Improving Neural Speech Synthesis Through Linear Prediction"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6501
| null |
Validated
| null | null |
null |
{
"abstract": " We theoretically investigate a scheme in which backward coherent anti-Stokes\nRaman scattering (CARS) is significantly enhanced by using slow light.\nSpecifically, we reduce the group velocity of the Stokes excitation pulse by\nintroducing a coupling laser that causes electromagnetically induced\ntransparency (EIT). When the Stokes pulse has a spatial length shorter than the\nCARS wavelength, the backward CARS emission is significantly enhanced. We also\ninvestigated the possibility of applying this scheme as a CARS lidar with O2 or\nN2 as the EIT medium. We found that if nanosecond laser with large pulse energy\n(>1 J) and a telescope with large aperture (~10 m) are equipped in the lidar\nsystem, a CARS lidar could become much more sensitive than a spontaneous Raman\nlidar.\n",
"title": "Coherent anti-Stokes Raman Scattering Lidar Using Slow Light: A Theoretical Study"
}
| null | null | null | null | true | null |
6502
| null |
Default
| null | null |
null |
{
"abstract": " A survey of goodness-of-fit and symmetry tests based on the characterization\nproperties of distributions is presented. This approach became popular in\nrecent years. In most cases the test statistics are functionals of\n$U$-empirical processes. The limiting distributions and large deviations of new\nstatistics under the null hypothesis are described. Their local Bahadur\nefficiency for various parametric alternatives is calculated and compared with\neach other as well as with diverse previously known tests. We also describe new\ndirections of possible research in this domain.\n",
"title": "Tests based on characterizations, and their efficiencies: a survey"
}
| null | null | null | null | true | null |
6503
| null |
Default
| null | null |
null |
{
"abstract": " In this article we introduce how to put vague hyperprior on Dirichlet\ndistribution, and we update the parameter of it by adaptive rejection sampling\n(ARS). Finally we analyze this hyperprior in an over-fitted mixture model by\nsome synthetic experiments.\n",
"title": "Hyperprior on symmetric Dirichlet distribution"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6504
| null |
Validated
| null | null |
null |
{
"abstract": " This is a comment to the paper 'A study of problems encountered in Granger\ncausality analysis from a neuroscience perspective'. We agree that\ninterpretation issues of Granger Causality in Neuroscience exist (partially due\nto the historical unfortunate use of the name 'causality', as nicely described\nin previous literature). On the other hand we think that the paper uses a\nformulation of Granger causality which is outdated (albeit still used), and in\ndoing so it dismisses the measure based on a suboptimal use of it. Furthermore,\nsince data from simulated systems are used, the pitfalls that are found with\nthe used formulation are intended to be general, and not limited to\nneuroscience. It would be a pity if this paper, even written in good faith,\nbecame a wildcard against all possible applications of Granger Causality,\nregardless of the hard work of colleagues aiming to seriously address the\nmethodological and interpretation pitfalls. In order to provide a balanced\nview, we replicated their simulations used the updated State Space\nimplementation, proposed already some years ago, in which the pitfalls are\nmitigated or directly solved.\n",
"title": "On the interpretability and computational reliability of frequency-domain Granger causality"
}
| null | null | null | null | true | null |
6505
| null |
Default
| null | null |
null |
{
"abstract": " Topological matter is a popular topic in both condensed matter and cold atom\nresearch. In the past decades, a variety of models have been identified with\nfascinating topological features. Some, but not all, of the models can be found\nin materials. As a fully controllable system, cold atoms trapped in optical\nlattices provide an ideal platform to simulate and realize these topological\nmodels. Here we present a proposal for synthesizing topological models in cold\natoms based on a one-dimensional (1D) spin-dependent optical lattice potential.\nIn our system, features such as staggered tunneling, staggered Zeeman field,\nnearest-neighbor interaction, beyond-near-neighbor tunneling, etc. can be\nreadily realized. They underlie the emergence of various topological phases.\nOur proposal can be realized with current technology and hence has potential\napplications in quantum simulation of topological matter.\n",
"title": "Artificial topological models based on a one-dimensional spin-dependent optical lattice"
}
| null | null |
[
"Physics"
] | null | true | null |
6506
| null |
Validated
| null | null |
null |
{
"abstract": " We discuss the Ramsey property, the existence of a stationary independence\nrelation and the coherent extension property for partial isometries (coherent\nEPPA) for all classes of metrically homogeneous graphs from Cherlin's\ncatalogue, which is conjectured to include all such structures. We show that,\nwith the exception of tree-like graphs, all metric spaces in the catalogue have\nprecompact Ramsey expansions (or lifts) with the expansion property. With two\nexceptions we can also characterise the existence of a stationary independence\nrelation and the coherent EPPA.\nOur results can be seen as a new contribution to Nešetřil's\nclassification programme of Ramsey classes and as empirical evidence of the\nrecent convergence in techniques employed to establish the Ramsey property, the\nexpansion (or lift or ordering) property, EPPA and the existence of a\nstationary independence relation. At the heart of our proof is a canonical way\nof completing edge-labelled graphs to metric spaces in Cherlin's classes. The\nexistence of such a \"completion algorithm\" then allows us to apply several\nstrong results in the areas that imply EPPA and respectively the Ramsey\nproperty.\nThe main results have numerous corollaries on the automorphism groups of the\nFraïssé limits of the classes, such as amenability, unique ergodicity,\nexistence of universal minimal flows, ample generics, small index property,\n21-Bergman property and Serre's property (FA).\n",
"title": "Ramsey expansions of metrically homogeneous graphs"
}
| null | null | null | null | true | null |
6507
| null |
Default
| null | null |
null |
{
"abstract": " The existence of massive ($10^{11}$ solar masses) elliptical galaxies by\nredshift z~4 (when the Universe was 1.5 billion years old) necessitates the\npresence of galaxies with star-formation rates exceeding 100 solar masses per\nyear at z>6 (corresponding to an age of the Universe of less than 1 billion\nyears). Surveys have discovered hundreds of galaxies at these early cosmic\nepochs, but their star-formation rates are more than an order of magnitude\nlower. The only known galaxies with very high star-formation rates at z>6 are,\nwith only one exception, the host galaxies of quasars, but these galaxies also\nhost accreting supermassive (more than $10^9$ solar masses) black holes, which\nprobably affect the properties of the galaxies. Here we report observations of\nan emission line of singly ionized carbon ([CII] at a wavelength of 158\nmicrometres) in four galaxies at z>6 that are companions of quasars, with\nvelocity offsets of less than 600 kilometers per second and linear offsets of\nless than 600 kiloparsecs. The discovery of these four galaxies was\nserendipitous; they are close to their companion quasars and appear bright in\nthe far-infrared. On the basis of the [CII] measurements, we estimate\nstar-formation rates in the companions of more than 100 solar masses per year.\nThese sources are similar to the host galaxies of the quasars in [CII]\nbrightness, linewidth and implied dynamical masses, but do not show evidence\nfor accreting supermassive black holes. Similar systems have previously been\nfound at lower redshift. We find such close companions in four out of\ntwenty-five z>6 quasars surveyed, a fraction that needs to be accounted for in\nsimulations. If they are representative of the bright end of the [CII]\nluminosity function, then they can account for the population of massive\nelliptical galaxies at z~4 in terms of cosmic space density.\n",
"title": "Rapidly star-forming galaxies adjacent to quasars at redshifts exceeding 6"
}
| null | null |
[
"Physics"
] | null | true | null |
6508
| null |
Validated
| null | null |
null |
{
"abstract": " A parametrization of irreducible unitary representations associated with the\nregular adjoint orbits of a hyperspecial compact subgroup of a reductive group\nover a non-dyadic non-archimedean local filed is presented. The parametrization\nis given by means of (a subset of) the character group of certain finite\nabelian groups arising from the reductive group. Our method is based upon\nCliffod's theory and Weil representations over finite fields. It works under an\nassumption of the triviality of certain Schur multipliers defined for an\nalgebraic group over a finite field. The assumption of the triviality has good\nevidences in the case of general linear groups and highly probable in general.\n",
"title": "Regular irreducible characters of a hyperspecial compact group"
}
| null | null | null | null | true | null |
6509
| null |
Default
| null | null |
null |
{
"abstract": " In this work we study the problem of exploring surfaces and building compact\n3D representations of the environment surrounding a robot through active\nperception. We propose an online probabilistic framework that merges visual and\ntactile measurements using Gaussian Random Field and Gaussian Process Implicit\nSurfaces. The system investigates incomplete point clouds in order to find a\nsmall set of regions of interest which are then physically explored with a\nrobotic arm equipped with tactile sensors. We show experimental results\nobtained using a PrimeSense camera, a Kinova Jaco2 robotic arm and Optoforce\nsensors on different scenarios. We then demonstrate how to use the online\nframework for object detection and terrain classification.\n",
"title": "Active Exploration Using Gaussian Random Fields and Gaussian Process Implicit Surfaces"
}
| null | null | null | null | true | null |
6510
| null |
Default
| null | null |
null |
{
"abstract": " Prosociality is fundamental to human social life, and, accordingly, much\nresearch has attempted to explain human prosocial behavior. Capraro and Rand\n(Judgment and Decision Making, 13, 99-111, 2018) recently provided experimental\nevidence that prosociality in anonymous, one-shot interactions (such as\nPrisoner's Dilemma and Dictator Game experiments) is not driven by\noutcome-based social preferences - as classically assumed - but by a\ngeneralized morality preference for \"doing the right thing\". Here we argue that\nthe key experiments reported in Capraro and Rand (2018) comprise prominent\nmethodological confounds and open questions that bear on influential\npsychological theory. Specifically, their design confounds: (i) preferences for\nefficiency with self-interest; and (ii) preferences for action with preferences\nfor morality. Furthermore, their design fails to dissociate the preference to\ndo \"good\" from the preference to avoid doing \"bad\". We thus designed and\nconducted a preregistered, refined and extended test of the morality preference\nhypothesis (N=801). Consistent with this hypothesis, our findings indicate that\nprosociality in the anonymous, one-shot Dictator Game is driven by preferences\nfor doing the morally right thing. Inconsistent with influential psychological\ntheory, however, our results suggest the preference to do \"good\" was as potent\nas the preference to avoid doing \"bad\" in this case.\n",
"title": "Doing good vs. avoiding bad in prosocial choice: A refined test and extension of the morality preference hypothesis"
}
| null | null | null | null | true | null |
6511
| null |
Default
| null | null |
null |
{
"abstract": " Territorial control is a key aspect shaping the dynamics of civil war.\nDespite its importance, we lack data on territorial control that are\nfine-grained enough to account for subnational spatio-temporal variation and\nthat cover a large set of conflicts. To resolve this issue, we propose a\ntheoretical model of the relationship between territorial control and tactical\nchoice in civil war and outline how Hidden Markov Models (HMMs) are suitable to\ncapture theoretical intuitions and estimate levels of territorial control. We\ndiscuss challenges of using HMMs in this application and mitigation strategies\nfor future work.\n",
"title": "Measuring Territorial Control in Civil Wars Using Hidden Markov Models: A Data Informatics-Based Approach"
}
| null | null | null | null | true | null |
6512
| null |
Default
| null | null |
null |
{
"abstract": " We construct a virtual quandle for links in lens spaces $L(p,q)$, with $q=1$.\nThis invariant has two valuable advantages over an ordinary fundamental quandle\nfor links in lens spaces: the virtual quandle is an essential invariant and the\npresentation of the virtual quandle can be easily written from the band diagram\nof a link.\n",
"title": "Virtual quandle for links in lens spaces"
}
| null | null | null | null | true | null |
6513
| null |
Default
| null | null |
null |
{
"abstract": " We derive an online learning algorithm with improved regret guarantees for\n`easy' loss sequences. We consider two types of `easiness': (a) stochastic loss\nsequences and (b) adversarial loss sequences with small effective range of the\nlosses. While a number of algorithms have been proposed for exploiting small\neffective range in the full information setting, Gerchinovitz and Lattimore\n[2016] have shown the impossibility of regret scaling with the effective range\nof the losses in the bandit setting. We show that just one additional\nobservation per round is sufficient to circumvent the impossibility result. The\nproposed Second Order Difference Adjustments (SODA) algorithm requires no prior\nknowledge of the effective range of the losses, $\\varepsilon$, and achieves an\n$O(\\varepsilon \\sqrt{KT \\ln K}) + \\tilde{O}(\\varepsilon K \\sqrt[4]{T})$\nexpected regret guarantee, where $T$ is the time horizon and $K$ is the number\nof actions. The scaling with the effective loss range is achieved under\nsignificantly weaker assumptions than those made by Cesa-Bianchi and Shamir\n[2018] in an earlier attempt to circumvent the impossibility result. We also\nprovide a regret lower bound of $\\Omega(\\varepsilon\\sqrt{T K})$, which almost\nmatches the upper bound. In addition, we show that in the stochastic setting\nSODA achieves an $O\\left(\\sum_{a:\\Delta_a>0}\n\\frac{K\\varepsilon^2}{\\Delta_a}\\right)$ pseudo-regret bound that holds\nsimultaneously with the adversarial regret guarantee. In other words, SODA is\nsafe against an unrestricted oblivious adversary and provides improved regret\nguarantees for at least two different types of `easiness' simultaneously.\n",
"title": "Adaptation to Easy Data in Prediction with Limited Advice"
}
| null | null | null | null | true | null |
6514
| null |
Default
| null | null |
null |
{
"abstract": " After a discussion of the Frauchiger-Renner argument that no 'single- world'\ninterpretation of quantum mechanics can be self-consistent, I propose a\n'Bohrian' alternative to many-worlds or QBism as the rational option.\n",
"title": "Why Bohr was (Mostly) Right"
}
| null | null | null | null | true | null |
6515
| null |
Default
| null | null |
null |
{
"abstract": " We propose three measures of mutual dependence between multiple random\nvectors. All the measures are zero if and only if the random vectors are\nmutually independent. The first measure generalizes distance covariance from\npairwise dependence to mutual dependence, while the other two measures are sums\nof squared distance covariance. All the measures share similar properties and\nasymptotic distributions to distance covariance, and capture non-linear and\nnon-monotone mutual dependence between the random vectors. Inspired by complete\nand incomplete V-statistics, we define the empirical measures and simplified\nempirical measures as a trade-off between the complexity and power when testing\nmutual independence. Implementation of the tests is demonstrated by both\nsimulation results and real data examples.\n",
"title": "Generalizing Distance Covariance to Measure and Test Multivariate Mutual Dependence"
}
| null | null | null | null | true | null |
6516
| null |
Default
| null | null |
null |
{
"abstract": " Let $X_1, \\ldots, X_n\\in\\mathbb{R}^p$ be i.i.d. random vectors. We aim to\nperform simultaneous inference for the mean vector $\\mathbb{E} (X_i)$ with\nfinite polynomial moments and an ultra high dimension. Our approach is based on\nthe truncated sample mean vector. A Gaussian approximation result is derived\nfor the latter under the very mild finite polynomial ($(2+\\theta)$-th) moment\ncondition and the dimension $p$ can be allowed to grow exponentially with the\nsample size $n$. Based on this result, we propose an innovative resampling\nmethod to construct simultaneous confidence intervals for mean vectors.\n",
"title": "Simultaneous Inference for High Dimensional Mean Vectors"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
6517
| null |
Validated
| null | null |
null |
{
"abstract": " We consider optimal/efficient power allocation policies in a single/multihop\nwireless network in the presence of hard end-to-end deadline delay constraints\non the transmitted packets. Such constraints can be useful for real time voice\nand video. Power is consumed in only transmission of the data. We consider the\ncase when the power used in transmission is a convex function of the data\ntransmitted. We develop a computationally efficient online algorithm, which\nminimizes the average power for the single hop. We model this problem as\ndynamic program (DP) and obtain the optimal solution. Next, we generalize it to\nthe multiuser, multihop scenario when there are multiple real time streams with\ndifferent hard deadline constraints.\n",
"title": "Joint Routing, Scheduling and Power Control Providing Hard Deadline in Wireless Multihop Networks"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6518
| null |
Validated
| null | null |
null |
{
"abstract": " Action potentials are the basic unit of information in the nervous system and\ntheir reliable detection and decoding holds the key to understanding how the\nbrain generates complex thought and behavior. Transducing these signals into\nmicrowave field oscillations can enable wireless sensors that report on brain\nactivity through magnetic induction. In the present work we demonstrate that\naction potentials from crayfish lateral giant neuron can trigger microwave\noscillations in spin-torque nano-oscillators. These nanoscale devices take as\ninput small currents and convert them to microwave current oscillations that\ncan wirelessly broadcast neuronal activity, opening up the possibility for\ncompact neuro-sensors. We show that action potentials activate microwave\noscillations in spin-torque nano-oscillators with an amplitude that follows the\naction potential signal, demonstrating that the device has both the sensitivity\nand temporal resolution to respond to action potentials from a single neuron.\nThe activation of magnetic oscillations by action potentials, together with the\nsmall footprint and the high frequency tunability, makes these devices\npromising candidates for high resolution sensing of bioelectric signals from\nneural tissues. These device attributes may be useful for design of\nhigh-throughput bi-directional brain-machine interfaces.\n",
"title": "Activation of Microwave Fields in a Spin-Torque Nano-Oscillator by Neuronal Action Potentials"
}
| null | null | null | null | true | null |
6519
| null |
Default
| null | null |
null |
{
"abstract": " We introduce a rigorous definition of general power-spectrum responses as\nresummed vertices with two hard and $n$ soft momenta in cosmological\nperturbation theory. These responses measure the impact of long-wavelength\nperturbations on the local small-scale power spectrum. The kinematic structure\nof the responses (i.e., their angular dependence) can be decomposed\nunambiguously through a \"bias\" expansion of the local power spectrum, with a\nfixed number of physical response coefficients, which are only a function of\nthe hard wavenumber $k$. Further, the responses up to $n$-th order completely\ndescribe the $(n+2)$-point function in the squeezed limit, i.e. with two hard\nand $n$ soft modes, which one can use to derive the response coefficients. This\ngeneralizes previous results, which relate the angle-averaged squeezed limit to\nisotropic response coefficients. We derive the complete expression of first-\nand second-order responses at leading order in perturbation theory, and present\nextrapolations to nonlinear scales based on simulation measurements of the\nisotropic response coefficients. As an application, we use these results to\npredict the non-Gaussian part of the angle-averaged matter power spectrum\ncovariance ${\\rm Cov}^{\\rm NG}_{\\ell = 0}(k_1,k_2)$, in the limit where one of\nthe modes, say $k_2$, is much smaller than the other. Without any free\nparameters, our model results are in very good agreement with simulations for\n$k_2 \\lesssim 0.06\\ h/{\\rm Mpc}$, and for any $k_1 \\gtrsim 2 k_2$. The\nwell-defined kinematic structure of the power spectrum response also permits a\nquick evaluation of the angular dependence of the covariance matrix. While we\nfocus on the matter density field, the formalism presented here can be\ngeneralized to generic tracers such as galaxies.\n",
"title": "Responses in Large-Scale Structure"
}
| null | null | null | null | true | null |
6520
| null |
Default
| null | null |
null |
{
"abstract": " Let us be given two graphs $\\Gamma_1$, $\\Gamma_2$ of $n$ vertices. Are they\nisomorphic? If they are, the set of isomorphisms from $\\Gamma_1$ to $\\Gamma_2$\ncan be identified with a coset $H\\cdot\\pi$ inside the symmetric group on $n$\nelements. How do we find $\\pi$ and a set of generators of $H$?\nThe challenge of giving an always efficient algorithm answering these\nquestions remained open for a long time. Babai has recently shown how to solve\nthese problems -- and others linked to them -- in quasi-polynomial time, i.e.\nin time $\\exp\\left(O(\\log n)^{O(1)}\\right)$. His strategy is based in part on\nthe algorithm by Luks (1980/82), who solved the case of graphs of bounded\ndegree.\n",
"title": "Graph isomorphisms in quasi-polynomial time"
}
| null | null | null | null | true | null |
6521
| null |
Default
| null | null |
null |
{
"abstract": " Recent breakthroughs in Deep Learning (DL) applications have made DL models a\nkey component in almost every modern computing system. The increased popularity\nof DL applications deployed on a wide-spectrum of platforms have resulted in a\nplethora of design challenges related to the constraints introduced by the\nhardware itself. What is the latency or energy cost for an inference made by a\nDeep Neural Network (DNN)? Is it possible to predict this latency or energy\nconsumption before a model is trained? If yes, how can machine learners take\nadvantage of these models to design the hardware-optimal DNN for deployment?\nFrom lengthening battery life of mobile devices to reducing the runtime\nrequirements of DL models executing in the cloud, the answers to these\nquestions have drawn significant attention.\nOne cannot optimize what isn't properly modeled. Therefore, it is important\nto understand the hardware efficiency of DL models during serving for making an\ninference, before even training the model. This key observation has motivated\nthe use of predictive models to capture the hardware performance or energy\nefficiency of DL applications. Furthermore, DL practitioners are challenged\nwith the task of designing the DNN model, i.e., of tuning the hyper-parameters\nof the DNN architecture, while optimizing for both accuracy of the DL model and\nits hardware efficiency. Therefore, state-of-the-art methodologies have\nproposed hardware-aware hyper-parameter optimization techniques. In this paper,\nwe provide a comprehensive assessment of state-of-the-art work and selected\nresults on the hardware-aware modeling and optimization for DL applications. We\nalso highlight several open questions that are poised to give rise to novel\nhardware-aware designs in the next few years, as DL applications continue to\nsignificantly impact associated hardware systems and platforms.\n",
"title": "Hardware-Aware Machine Learning: Modeling and Optimization"
}
| null | null | null | null | true | null |
6522
| null |
Default
| null | null |
null |
{
"abstract": " The Erd\\H os unit distance conjecture in the plane says that the number of\npairs of points from a point set of size $n$ separated by a fixed (Euclidean)\ndistance is $\\leq C_{\\epsilon} n^{1+\\epsilon}$ for any $\\epsilon>0$. The best\nknown bound is $Cn^{\\frac{4}{3}}$. We show that if the set under consideration\nis well-distributed and the fixed distance is much smaller than the diameter of\nthe set, then the exponent $\\frac{4}{3}$ is significantly improved.\nCorresponding results are also established in higher dimensions. The results\nare obtained by solving the corresponding continuous problem and using a\ncontinuous-to-discrete conversion mechanism. The degree of sharpness of results\nis tested using the known results on the distribution of lattice points dilates\nof convex domains.\nWe also introduce the following variant of the Erd\\H os unit distance\nproblem: how many pairs of points from a set of size $n$ are separated by an\ninteger distance? We obtain some results in this direction and formulate a\nconjecture.\n",
"title": "On the unit distance problem"
}
| null | null | null | null | true | null |
6523
| null |
Default
| null | null |
null |
{
"abstract": " We study the ground state of the 1D Kitaev-Heisenberg (KH) model using the\ndensity-matrix renormalization group and Lanczos exact diagonalization methods.\nWe obtain a rich ground-state phase diagram as a function of the ratio between\nHeisenberg ($J=\\cos\\phi)$ and Kitaev ($K=\\sin\\phi$) interactions. Depending on\nthe ratio, the system exhibits four long-range ordered states:\nferromagnetic-$z$ , ferromagnetic-$xy$, staggered-$xy$, Néel-$z$, and two\nliquid states: Tomonaga-Luttinger liquid and spiral-$xy$. The two Kitaev points\n$\\phi=\\frac{\\pi}{2}$ and $\\phi=\\frac{3\\pi}{2}$ are singular. The\n$\\phi$-dependent phase diagram is similar to that for the 2D honeycomb-lattice\nKH model. Remarkably, all the ordered states of the honeycomb-lattice KH model\ncan be interpreted in terms of the coupled KH chains. We also discuss the\nmagnetic structure of the K-intercalated RuCl$_3$, a potential Kitaev material,\nin the framework of the 1D KH model. Furthermore, we demonstrate that the\nlow-lying excitations of the 1D KH Hamiltonian can be explained within the\ncombination of the known six-vertex model and spin-wave theory.\n",
"title": "Ordered states in the Kitaev-Heisenberg model: From 1D chains to 2D honeycomb"
}
| null | null | null | null | true | null |
6524
| null |
Default
| null | null |
null |
{
"abstract": " We study quadratic functionals on $L^2(\\mathbb{R}^d)$ that generate seminorms\nin the fractional Sobolev space $H^s(\\mathbb{R}^d)$ for $0 < s < 1$. The\nfunctionals under consideration appear in the study of Markov jump processes\nand, independently, in recent research on the Boltzmann equation. The\nfunctional measures differentiability of a function $f$ in a similar way as the\nseminorm of $H^s(\\mathbb{R}^d)$. The major difference is that differences $f(y)\n- f(x)$ are taken into account only if $y$ lies in some double cone with apex\nat $x$ or vice versa. The configuration of double cones is allowed to be\ninhomogeneous without any assumption on the spatial regularity. We prove that\nthe resulting seminorm is comparable to the standard one of\n$H^s(\\mathbb{R}^d)$. The proof follows from a similar result on discrete\nquadratic forms in $\\mathbb{Z}^d$, which is our second main result. We\nestablish a general scheme for discrete approximations of nonlocal quadratic\nforms. Applications to Markov jump processes are discussed.\n",
"title": "Quadratic forms and Sobolev spaces of fractional order"
}
| null | null | null | null | true | null |
6525
| null |
Default
| null | null |
null |
{
"abstract": " This paper introduces a general Bayesian non- parametric latent feature model\nsuitable to per- form automatic exploratory analysis of heterogeneous datasets,\nwhere the attributes describing each object can be either discrete, continuous\nor mixed variables. The proposed model presents several important properties.\nFirst, it accounts for heterogeneous data while can be inferred in linear time\nwith respect to the number of objects and attributes. Second, its Bayesian\nnonparametric nature allows us to automatically infer the model complexity from\nthe data, i.e., the number of features necessary to capture the latent\nstructure in the data. Third, the latent features in the model are\nbinary-valued variables, easing the interpretability of the obtained latent\nfeatures in data exploration tasks.\n",
"title": "General Latent Feature Modeling for Data Exploration Tasks"
}
| null | null | null | null | true | null |
6526
| null |
Default
| null | null |
null |
{
"abstract": " Most massive stars form in dense clusters where gravitational interactions\nwith other stars may be common. The two nearest forming massive stars, the BN\nobject and Source I, located behind the Orion Nebula, were ejected with\nvelocities of $\\sim$29 and $\\sim$13 km s$^{-1}$ about 500 years ago by such\ninteractions. This event generated an explosion in the gas. New ALMA\nobservations show in unprecedented detail, a roughly spherically symmetric\ndistribution of over a hundred $^{12}$CO J=2$-$1 streamers with velocities\nextending from V$_{LSR}$ =$-$150 to +145 km s$^{-1}$. The streamer radial\nvelocities increase (or decrease) linearly with projected distance from the\nexplosion center, forming a `Hubble Flow' confined to within 50 arcseconds of\nthe explosion center. They point toward the high proper-motion, shock-excited\nH$_2$ and [Fe ii ] `fingertips' and lower-velocity CO in the H$_2$ wakes\ncomprising Orion's `fingers'. In some directions, the H$_2$ `fingers' extend\nmore than a factor of two farther from the ejection center than the CO\nstreamers. Such deviations from spherical symmetry may be caused by ejecta\nrunning into dense gas or the dynamics of the N-body interaction that ejected\nthe stars and produced the explosion. This $\\sim$10$^{48}$ erg event may have\nbeen powered by the release of gravitational potential energy associated with\nthe formation of a compact binary or a protostellar merger. Orion may be the\nprototype for a new class of stellar explosion responsible for luminous\ninfrared transients in nearby galaxies.\n",
"title": "The ALMA View of the OMC1 Explosion in Orion"
}
| null | null |
[
"Physics"
] | null | true | null |
6527
| null |
Validated
| null | null |
null |
{
"abstract": " Electroencephalography (EEG) is an extensively-used and well-studied\ntechnique in the field of medical diagnostics and treatment for brain\ndisorders, including epilepsy, migraines, and tumors. The analysis and\ninterpretation of EEGs require physicians to have specialized training, which\nis not common even among most doctors in the developed world, let alone the\ndeveloping world where physician shortages plague society. This problem can be\naddressed by teleEEG that uses remote EEG analysis by experts or by local\ncomputer processing of EEGs. However, both of these options are prohibitively\nexpensive and the second option requires abundant computing resources and\ninfrastructure, which is another concern in developing countries where there\nare resource constraints on capital and computing infrastructure. In this work,\nwe present a cloud-based deep neural network approach to provide decision\nsupport for non-specialist physicians in EEG analysis and interpretation. Named\n`neurology-as-a-service,' the approach requires almost no manual intervention\nin feature engineering and in the selection of an optimal architecture and\nhyperparameters of the neural network. In this study, we deploy a pipeline that\nincludes moving EEG data to the cloud and getting optimal models for various\nclassification tasks. Our initial prototype has been tested only in developed\nworld environments to-date, but our intention is to test it in developing world\nenvironments in future work. We demonstrate the performance of our proposed\napproach using the BCI2000 EEG MMI dataset, on which our service attains 63.4%\naccuracy for the task of classifying real vs. imaginary activity performed by\nthe subject, which is significantly higher than what is obtained with a shallow\napproach such as support vector machines.\n",
"title": "Neurology-as-a-Service for the Developing World"
}
| null | null | null | null | true | null |
6528
| null |
Default
| null | null |
null |
{
"abstract": " Consider the problem of estimating a low-rank symmetric matrix when its\nentries are perturbed by Gaussian noise, a setting that is known as `spiked\nmodel' or `deformed Wigner matrix'. If the empirical distribution of the\nentries of the spikes is known, optimal estimators that exploit this knowledge\ncan substantially outperform spectral approaches. Recent work characterizes the\naccuracy of Bayes-optimal estimators in the high-dimensional limit. In this\npaper we present a practical algorithm that can achieve Bayes-optimal accuracy\nabove the spectral threshold. A bold conjecture from statistical physics posits\nthat no polynomial-time algorithm achieves optimal error below the same\nthreshold (unless the best estimator is trivial). Our approach uses Approximate\nMessage Passing (AMP) in conjunction with a spectral initialization. AMP has\nproven successful in a variety of statistical problem, and are amenable to\nexact asymptotic analysis via state evolution. Unfortunately, state evolution\nis uninformative when the algorithm is initialized near an unstable fixed\npoint, as it often happens in matrix estimation. We develop a new analysis of\nAMP that allows for spectral initializations, and builds on a decoupling\nbetween the outlier eigenvectors and the bulk in the spiked random matrix\nmodel. Our main theorem is general and applies beyond matrix estimation.\nHowever, we use it to derive detailed predictions for the problem of estimating\na rank-one matrix in noise. Special cases of these problem are closely related\n-via universality arguments- to the network community detection problem for two\nasymmetric communities. For general rank-one models, we show that AMP can be\nused to construct asymptotically valid confidence intervals. As a further\nillustration, we consider the example of a block-constant low-rank matrix with\nsymmetric blocks, which we refer to as `Gaussian Block Model'.\n",
"title": "Estimation of Low-Rank Matrices via Approximate Message Passing"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
6529
| null |
Validated
| null | null |
null |
{
"abstract": " Recent developments in quaternion-valued widely linear processing have\nestablished that the exploitation of complete second-order statistics requires\nconsideration of both the standard covariance and the three complementary\ncovariance matrices. Although such matrices have a tremendous amount of\nstructure and their decomposition is a powerful tool in a variety of\napplications, the non-commutative nature of the quaternion product has been\nprohibitive to the development of quaternion uncorrelating transforms. To this\nend, we introduce novel techniques for a simultaneous decomposition of the\ncovariance and complementary covariance matrices in the quaternion domain,\nwhereby the quaternion version of the Takagi factorisation is explored to\ndiagonalise symmetric quaternion-valued matrices. This gives new insights into\nthe quaternion uncorrelating transform (QUT) and forms a basis for the proposed\nquaternion approximate uncorrelating transform (QAUT) which simultaneously\ndiagonalises all four covariance matrices associated with improper quaternion\nsignals. The effectiveness of the proposed uncorrelating transforms is\nvalidated by simulations on both synthetic and real-world quaternion-valued\nsignals.\n",
"title": "Simultaneous diagonalisation of the covariance and complementary covariance matrices in quaternion widely linear signal processing"
}
| null | null | null | null | true | null |
6530
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we proved an arithmetic Siegel-Weil formula and the modularity\nof some arithmetic theta function on the modular curve $X_0(N)$ when $N$ is\nsquare free. In the process, we also construct some generalized Delta function\nfor $\\Gamma_0(N)$ and proved some explicit Kronecker limit formula for\nEisenstein series on $X_0(N)$.\n",
"title": "Arithmetic Siegel-Weil formula on $X_{0}(N)$"
}
| null | null | null | null | true | null |
6531
| null |
Default
| null | null |
null |
{
"abstract": " Motivated by $\\alpha$-attractor models, in this paper we consider a\nGauss-Bonnet inflation with E-model type of potential. We consider the\nGauss-Bonnet coupling function to be the same as the E-model potential. In the\nsmall $\\alpha$ limit we obtain an attractor at $r=0$ as expected, and in the\nlarge $\\alpha$ limit we recover the Gauss-Bonnet model with potential and\ncoupling function of the form $\\phi^{2n}$. We study perturbations and\nnon-Gaussianity in this setup and we find some constraints on the model's\nparameters in comparison with PLANCK datasets. We study also the reheating\nepoch after inflation in this setup. For this purpose, we seek the number of\ne-folds and temperature during reheating epoch. These quantities depend on the\nmodel's parameter and the effective equation of state of the dominating energy\ndensity in the reheating era. We find some observational constraints on these\nparameters.\n",
"title": "Perturbation, Non-Gaussianity and Reheating in a GB-$α$-Attractor Model"
}
| null | null |
[
"Physics"
] | null | true | null |
6532
| null |
Validated
| null | null |
null |
{
"abstract": " Magnetic nanoparticles are promising systems for biomedical applications and\nin particular for Magnetic Fluid Hyperthermia, a promising therapy that\nutilizes the heat released by such systems to damage tumor cells. We present an\nexperimental study of the physical properties that influences the capability of\nheat release, i.e. the Specific Loss Power, SLP, of three biocompatible\nferrofluid samples having a magnetic core of maghemite with different core\ndiameter d= 10.2, 14.6 and 19.7 nm. The SLP was measured as a function of\nfrequency f and intensity of the applied alternating magnetic field H, and it\nturned out to depend on the core diameter, as expected. The results allowed us\nto highlight experimentally that the physical mechanism responsible for the\nheating is size-dependent and to establish, at applied constant frequency, the\nphenomenological functional relationship SLP=cH^x, with 2<x<3 for all samples.\nThe x-value depends on sample size and field frequency/ intensity, here chosen\nin the typical range of operating magnetic hyperthermia devices. For the\nsmallest sample, the effective relaxation time Teff=19.5 ns obtained from SLP\ndata is in agreement with the value estimated from magnetization data, thus\nconfirming the validity of the Linear Response Theory model for this system at\nproperly chosen field intensity and frequency.\n",
"title": "Experimental determination of the frequency and field dependence of Specific Loss Power in Magnetic Fluid Hyperthermia"
}
| null | null | null | null | true | null |
6533
| null |
Default
| null | null |
null |
{
"abstract": " For fast and energy-efficient deployment of trained deep neural networks on\nresource-constrained embedded hardware, each learned weight parameter should\nideally be represented and stored using a single bit. Error-rates usually\nincrease when this requirement is imposed. Here, we report large improvements\nin error rates on multiple datasets, for deep convolutional neural networks\ndeployed with 1-bit-per-weight. Using wide residual networks as our main\nbaseline, our approach simplifies existing methods that binarize weights by\napplying the sign function in training; we apply scaling factors for each layer\nwith constant unlearned values equal to the layer-specific standard deviations\nused for initialization. For CIFAR-10, CIFAR-100 and ImageNet, and models with\n1-bit-per-weight requiring less than 10 MB of parameter memory, we achieve\nerror rates of 3.9%, 18.5% and 26.0% / 8.5% (Top-1 / Top-5) respectively. We\nalso considered MNIST, SVHN and ImageNet32, achieving 1-bit-per-weight test\nresults of 0.27%, 1.9%, and 41.3% / 19.1% respectively. For CIFAR, our error\nrates halve previously reported values, and are within about 1% of our\nerror-rates for the same network with full-precision weights. For networks that\noverfit, we also show significant improvements in error rate by not learning\nbatch normalization scale and offset parameters. This applies to both full\nprecision and 1-bit-per-weight networks. Using a warm-restart learning-rate\nschedule, we found that training for 1-bit-per-weight is just as fast as\nfull-precision networks, with better accuracy than standard schedules, and\nachieved about 98%-99% of peak performance in just 62 training epochs for\nCIFAR-10/100. For full training code and trained models in MATLAB, Keras and\nPyTorch see this https URL .\n",
"title": "Training wide residual networks for deployment using a single bit for each weight"
}
| null | null | null | null | true | null |
6534
| null |
Default
| null | null |
null |
{
"abstract": " With the passage of time and indulgence in Information Technology, network\nmanagement has proved its significance and has become one of the most important\nand challenging task in today's era of information flow. Communication networks\nimplement a high level of sophistication in managing and flowing the data\nthrough secure channels, to make it almost impossible for data loss. That is\nwhy there are many proposed solution that are currently implemented in wide\nrange of network-based applications like social networks and finance\napplications. The objective of this research paper is to propose a very\nreliable method of data flow: Choose best path for traffic using SDN\napplication. This is an IP based method in which our SDN application implements\nprovision of best possible path by filtering the requests on base of their IP\norigin. We'll distinguish among source and will provide the data flow with\nlowest traffic path, thus resulting in providing us minimum chances of data\nloss. A request to access our test application will be generated from host and\nrequest from each host will be distinguished by our SDN application that will\nget number of active users for all available servers and will redirect the\nrequest to server with minimum traffic load. It will also destroy sessions of\ninactive users, resulting in maintaining a best responsive channel for data\nflow.\n",
"title": "IP Based Traffic Recovery: An Optimal Approach using SDN Application for Data Center Network"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6535
| null |
Validated
| null | null |
null |
{
"abstract": " Data shaping is a coding technique that has been proposed to increase the\nlifetime of flash memory devices. Several data shaping codes have been\ndescribed in recent work, including endurance codes and direct shaping codes\nfor structured data. In this paper, we study information-theoretic properties\nof a general class of data shaping codes and prove a separation theorem stating\nthat optimal data shaping can be achieved by the concatenation of optimal\nlossless compression with optimal endurance coding. We also determine the\nexpansion factor that minimizes the total wear cost. Finally, we analyze the\nperformance of direct shaping codes and establish a condition for their\noptimality.\n",
"title": "Performance of Optimal Data Shaping Codes"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6536
| null |
Validated
| null | null |
null |
{
"abstract": " Global integration of information in the brain results from complex\ninteractions of segregated brain networks. Identifying the most influential\nneuronal populations that efficiently bind these networks is a fundamental\nproblem of systems neuroscience. Here we apply optimal percolation theory and\npharmacogenetic interventions in-vivo to predict and subsequently target nodes\nthat are essential for global integration of a memory network in rodents. The\ntheory predicts that integration in the memory network is mediated by a set of\nlow-degree nodes located in the nucleus accumbens. This result is confirmed\nwith pharmacogenetic inactivation of the nucleus accumbens, which eliminates\nthe formation of the memory network, while inactivations of other brain areas\nleave the network intact. Thus, optimal percolation theory predicts essential\nnodes in brain networks. This could be used to identify targets of\ninterventions to modulate brain function.\n",
"title": "Finding influential nodes for integration in brain networks using optimal percolation theory"
}
| null | null | null | null | true | null |
6537
| null |
Default
| null | null |
null |
{
"abstract": " Music highlights are valuable contents for music services. Most methods\nfocused on low-level signal features. We propose a method for extracting\nhighlights using high-level features from convolutional recurrent attention\nnetworks (CRAN). CRAN utilizes convolution and recurrent layers for sequential\nlearning with an attention mechanism. The attention allows CRAN to capture\nsignificant snippets for distinguishing between genres, thus being used as a\nhigh-level feature. CRAN was evaluated on over 32,000 popular tracks in Korea\nfor two months. Experimental results show our method outperforms three baseline\nmethods through quantitative and qualitative evaluations. Also, we analyze the\neffects of attention and sequence information on performance.\n",
"title": "Automatic Music Highlight Extraction using Convolutional Recurrent Attention Networks"
}
| null | null | null | null | true | null |
6538
| null |
Default
| null | null |
null |
{
"abstract": " We study the computational tractability of PAC reinforcement learning with\nrich observations. We present new provably sample-efficient algorithms for\nenvironments with deterministic hidden state dynamics and stochastic rich\nobservations. These methods operate in an oracle model of computation --\naccessing policy and value function classes exclusively through standard\noptimization primitives -- and therefore represent computationally efficient\nalternatives to prior algorithms that require enumeration. With stochastic\nhidden state dynamics, we prove that the only known sample-efficient algorithm,\nOLIVE, cannot be implemented in the oracle model. We also present several\nexamples that illustrate fundamental challenges of tractable PAC reinforcement\nlearning in such general settings.\n",
"title": "On Oracle-Efficient PAC RL with Rich Observations"
}
| null | null | null | null | true | null |
6539
| null |
Default
| null | null |
null |
{
"abstract": " We show a noise-induced transition in Josephson junction with fundamental as\nwell as second harmonic. A periodically modulated multiplicative colored noise\ncan stabilize an unstable configuration in such a system. The stabilization of\nthe unstable configuration has been captured in the effective potential of the\nsystem obtained by integrating out the high-frequency components of the noise.\nThis is a classical approach to understand the stability of an unstable\nconfiguration due to the presence of such stochasticity in the system and our\nnumerical analysis confirms the prediction from the analytical calculation.\n",
"title": "Noise induced transition in Josephson junction with second harmonic"
}
| null | null | null | null | true | null |
6540
| null |
Default
| null | null |
null |
{
"abstract": " Principal component analysis (PCA) is fundamental to statistical machine\nlearning. It extracts latent principal factors that contribute to the most\nvariation of the data. When data are stored across multiple machines, however,\ncommunication cost can prohibit the computation of PCA in a central location\nand distributed algorithms for PCA are thus needed. This paper proposes and\nstudies a distributed PCA algorithm: each node machine computes the top $K$\neigenvectors and transmits them to the central server; the central server then\naggregates the information from all the node machines and conducts a PCA based\non the aggregated information. We investigate the bias and variance for the\nresulting distributed estimator of the top $K$ eigenvectors. In particular, we\nshow that for distributions with symmetric innovation, the empirical top\neigenspaces are unbiased and hence the distributed PCA is \"unbiased\". We derive\nthe rate of convergence for distributed PCA estimators, which depends\nexplicitly on the effective rank of covariance, eigen-gap, and the number of\nmachines. We show that when the number of machines is not unreasonably large,\nthe distributed PCA performs as well as the whole sample PCA, even without full\naccess of whole data. The theoretical results are verified by an extensive\nsimulation study. We also extend our analysis to the heterogeneous case where\nthe population covariance matrices are different across local machines but\nshare similar top eigen-structures.\n",
"title": "Distributed Estimation of Principal Eigenspaces"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
6541
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper, we characterize all the irreducible Darboux polynomials and\npolynomial first integrals of FitzHugh-Nagumo (F-N) system. The method of the\nweight homogeneous polynomials and the characteristic curves is widely used to\ngive a complete classification of Darboux polynomials of a system. However,\nthis method does not work for F-N system. Here by considering the Darboux\npolynomials of an assistant system associated to F-N system, we classified the\ninvariant algebraic surfaces of F-N system. Our results show that there is no\ninvariant algebraic surface of F-N system in the biological parameters region.\n",
"title": "Invariant algebraic surfaces of the FitzHugh-Nagumo system"
}
| null | null | null | null | true | null |
6542
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we analyze the local and global boundary rigidity problem for\ngeneral Riemannian manifolds with boundary $(M,g)$ whose boundary is strictly\nconvex. We show that the boundary distance function, i.e., $d_g|_{\\partial\nM\\times\\partial M}$, known over suitable open sets of $\\partial M$ determines\n$g$ in suitable corresponding open subsets of $M$, up to the natural\ndiffeomorphism invariance of the problem. We also show that if there is a\nfunction on $M$ with suitable convexity properties relative to $g$ then\n$d_g|_{\\partial M\\times\\partial M}$ determines $g$ globally in the sense that\nif $d_g|_{\\partial M\\times\\partial M}=d_{\\tilde g}|_{\\partial M\\times \\partial\nM}$ then there is a diffeomorphism $\\psi$ fixing $\\partial M$ (pointwise) such\nthat $g=\\psi^*\\tilde g$. This global assumption is satisfied, for instance, for\nthe distance function from a given point if the manifold has no focal points\n(from that point).\nWe also consider the lens rigidity problem. The lens relation measures the\npoint of exit from $M$ and the direction of exit of geodesics issued from the\nboundary and the length of the geodesic. The lens rigidity problem is whether\nwe can determine the metric up to isometry from the lens relation. We solve the\nlens rigidity problem under the same global assumption mentioned above. This\nshows, for instance, that manifolds with a strictly convex boundary and\nnon-positive sectional curvature are lens rigid.\nThe key tool is the analysis of the geodesic X-ray transform on 2-tensors,\ncorresponding to a metric $g$, in the normal gauge, such as normal coordinates\nrelative to a hypersurface, where one also needs to allow microlocal weights.\nThis is handled by refining and extending our earlier results in the solenoidal\ngauge.\n",
"title": "Local and global boundary rigidity and the geodesic X-ray transform in the normal gauge"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6543
| null |
Validated
| null | null |
null |
{
"abstract": " We study a variant of the source identification game with training data in\nwhich part of the training data is corrupted by an attacker. In the addressed\nscenario, the defender aims at deciding whether a test sequence has been drawn\naccording to a discrete memoryless source $X \\sim P_X$, whose statistics are\nknown to him through the observation of a training sequence generated by $X$.\nIn order to undermine the correct decision under the alternative hypothesis\nthat the test sequence has not been drawn from $X$, the attacker can modify a\nsequence produced by a source $Y \\sim P_Y$ up to a certain distortion, and\ncorrupt the training sequence either by adding some fake samples or by\nreplacing some samples with fake ones. We derive the unique rationalizable\nequilibrium of the two versions of the game in the asymptotic regime and by\nassuming that the defender bases its decision by relying only on the first\norder statistics of the test and the training sequences. By mimicking Stein's\nlemma, we derive the best achievable performance for the defender when the\nfirst type error probability is required to tend to zero exponentially fast\nwith an arbitrarily small, yet positive, error exponent. We then use such a\nresult to analyze the ultimate distinguishability of any two sources as a\nfunction of the allowed distortion and the fraction of corrupted samples\ninjected into the training sequence.\n",
"title": "Adversarial Source Identification Game with Corrupted Training"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6544
| null |
Validated
| null | null |
null |
{
"abstract": " We prove a convexity theorem for Hamiltonian torus actions on compact\ncosymplectic manifolds. We show that compact toric cosymplectic manifolds are\nmapping tori of equivariant symplectomorphisms of toric symplectic manifolds.\n",
"title": "Toric actions and convexity in cosymplectic geometry"
}
| null | null | null | null | true | null |
6545
| null |
Default
| null | null |
null |
{
"abstract": " Given an odd vector field $Q$ on a supermanifold $M$ and a $Q$-invariant\ndensity $\\mu$ on $M$, under certain compactness conditions on $Q$, the value of\nthe integral $\\int_{M}\\mu$ is determined by the value of $\\mu$ on any\nneighborhood of the vanishing locus $N$ of $Q$. We present a formula for the\nintegral in the case where $N$ is a subsupermanifold which is appropriately\nnon-degenerate with respect to $Q$.\nIn the process, we discuss the linear algebra necessary to express our result\nin a coordinate independent way. We also extend stationary phase approximation\nand the Morse-Bott Lemma to supermanifolds.\n",
"title": "Localization and Stationary Phase Approximation on Supermanifolds"
}
| null | null | null | null | true | null |
6546
| null |
Default
| null | null |
null |
{
"abstract": " Multicopters are becoming increasingly important in both civil and military\nfields. Currently, most multicopter propulsion systems are designed by\nexperience and trial-and-error experiments, which are costly and ineffective.\nThis paper proposes a simple and practical method to help designers find the\noptimal propulsion system according to the given design requirements. First,\nthe modeling methods for four basic components of the propulsion system\nincluding propellers, motors, electric speed controls, and batteries are\nstudied respectively. Secondly, the whole optimization design problem is\nsimplified and decoupled into several sub-problems. By solving these\nsub-problems, the optimal parameters of each component can be obtained\nrespectively. Finally, based on the obtained optimal component parameters, the\noptimal product of each component can be quickly located and determined from\nthe corresponding database. Experiments and statistical analyses demonstrate\nthe effectiveness of the proposed method.\n",
"title": "An Analytical Design Optimization Method for Electric Propulsion Systems of Multicopter UAVs with Desired Hovering Endurance"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6547
| null |
Validated
| null | null |
null |
{
"abstract": " We introduce a new technique for determining x-ray fluorescence line energies\nand widths, and we present measurements made with this technique of 22 x-ray L\nlines from lanthanide-series elements. The technique uses arrays of\ntransition-edge sensors, microcalorimeters with high energy-resolving power\nthat simultaneously observe both calibrated x-ray standards and the x-ray\nemission lines under study. The uncertainty in absolute line energies is\ngenerally less than 0.4 eV in the energy range of 4.5 keV to 7.5 keV. Of the\nseventeen line energies of neodymium, samarium, and holmium, thirteen are found\nto be consistent with the available x-ray reference data measured after 1990;\nonly two of the four lines for which reference data predate 1980, however, are\nconsistent with our results. Five lines of terbium are measured with\nuncertainties that improve on those of existing data by factors of two or more.\nThese results eliminate a significant discrepancy between measured and\ncalculated x-ray line energies for the terbium Ll line (5.551 keV). The line\nwidths are also measured, with uncertainties of 0.6 eV or less on the\nfull-width at half-maximum in most cases. These measurements were made with an\narray of approximately one hundred superconducting x- ray microcalorimeters,\neach sensitive to an energy band from 1 keV to 8 keV. No energy-dispersive\nspectrometer has previously been used for absolute-energy estimation at this\nlevel of accuracy. Future spectrometers, with superior linearity and energy\nresolution, will allow us to improve on these results and expand the\nmeasurements to more elements and a wider range of line energies.\n",
"title": "A Reassessment of Absolute Energies of the X-ray L Lines of Lanthanide Metals"
}
| null | null | null | null | true | null |
6548
| null |
Default
| null | null |
null |
{
"abstract": " The coupling of vocal fold (source) and vocal tract (filter) is one of the\nmost critical factors in source-filter articulation theory. The traditional\nlinear source-filter theory has been challenged by current research which\nclearly shows the impact of acoustic loading on the dynamic behavior of the\nvocal fold vibration as well as the variations in the glottal flow pulses\nshape. This paper outlines the underlying mechanism of source-filter\ninteractions; demonstrates the design and working principles of coupling for\nthe various existing vocal cord and vocal tract biomechanical models. For our\nstudy, we have considered self-oscillating lumped-element models of the\nacoustic source and computational models of the vocal tract as articulators. To\nunderstand the limitations of source-filter interactions which are associated\nwith each of those models, we compare them concerning their mechanical design,\nacoustic and physiological characteristics and aerodynamic simulation.\n",
"title": "Limitations of Source-Filter Coupling In Phonation"
}
| null | null | null | null | true | null |
6549
| null |
Default
| null | null |
null |
{
"abstract": " We are interested in the probability that two randomly selected neighbors of\na random vertex of degree (at least) $k$ are adjacent. We evaluate this\nprobability for a power law random intersection graph, where each vertex is\nprescribed a collection of attributes and two vertices are adjacent whenever\nthey share a common attribute. We show that the probability obeys the scaling\n$k^{-\\delta}$ as $k\\to+\\infty$. Our results are mathematically rigorous. The\nparameter $0\\le \\delta\\le 1$ is determined by the tail indices of power law\nrandom weights defining the links between vertices and attributes.\n",
"title": "Correlation between clustering and degree in affiliation networks"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6550
| null |
Validated
| null | null |
null |
{
"abstract": " Discourse connectives (e.g. however, because) are terms that can explicitly\nconvey a discourse relation within a text. While discourse connectives have\nbeen shown to be an effective clue to automatically identify discourse\nrelations, they are not always used to convey such relations, thus they should\nfirst be disambiguated between discourse-usage non-discourse-usage. In this\npaper, we investigate the applicability of features proposed for the\ndisambiguation of English discourse connectives for French. Our results with\nthe French Discourse Treebank (FDTB) show that syntactic and lexical features\ndeveloped for English texts are as effective for French and allow the\ndisambiguation of French discourse connectives with an accuracy of 94.2%.\n",
"title": "Automatic Disambiguation of French Discourse Connectives"
}
| null | null |
[
"Computer Science"
] | null | true | null |
6551
| null |
Validated
| null | null |
null |
{
"abstract": " Financial crime is a rampant but hidden threat. In spite of this, predictive\npolicing systems disproportionately target \"street crime\" rather than white\ncollar crime. This paper presents the White Collar Crime Early Warning System\n(WCCEWS), a white collar crime predictive model that uses random forest\nclassifiers to identify high risk zones for incidents of financial crime.\n",
"title": "Predicting Financial Crime: Augmenting the Predictive Policing Arsenal"
}
| null | null | null | null | true | null |
6552
| null |
Default
| null | null |
null |
{
"abstract": " We consider explicit polar constructions of blocklength $n\\rightarrow\\infty$\nfor the two extreme cases of code rates $R\\rightarrow1$ and $R\\rightarrow0.$\nFor code rates $R\\rightarrow1,$ we design codes with complexity order of $n\\log\nn$ in code construction, encoding, and decoding. These codes achieve the\nvanishing output bit error rates on the binary symmetric channels with any\ntransition error probability $p\\rightarrow 0$ and perform this task with a\nsubstantially smaller redundancy $(1-R)n$ than do other known high-rate codes,\nsuch as BCH codes or Reed-Muller (RM). We then extend our design to the\nlow-rate codes that achieve the vanishing output error rates with the same\ncomplexity order of $n\\log n$ and an asymptotically optimal code rate\n$R\\rightarrow0$ for the case of $p\\rightarrow1/2.$\n",
"title": "Polar codes with a stepped boundary"
}
| null | null | null | null | true | null |
6553
| null |
Default
| null | null |
null |
{
"abstract": " In the present article the classical problem of electromagnetic scattering by\na single homogeneous sphere is revisited. Main focus is the study of the\nscattering behavior as a function of the material contrast and the size\nparameters for all electric and magnetic resonances of a dielectric sphere.\nSpecifically, the Padé approximants are introduced and utilized as an\nalternative system expansion of the Mie coefficients. Low order Padé\napproximants can give compact and physically insightful expressions for the\nscattering system and the enabled dynamic mechanisms. Higher order approximants\nare used for predicting accurately the resonant pole spectrum. These results\nare summarized into general pole formulae, covering up to fifth order magnetic\nand forth order electric resonances of a small dielectric sphere. Additionally,\nthe connection between the radiative damping process and the resonant linewidth\nis investigated. The results obtained reveal the fundamental connection of the\nradiative damping mechanism with the maximum width occurring for each\nresonance. Finally, the suggested system ansatz is used for studying the\nresonant absorption maximum through a circuit-inspired perspective.\n",
"title": "Resonant Scattering Characteristics of Homogeneous Dielectric Sphere"
}
| null | null | null | null | true | null |
6554
| null |
Default
| null | null |
null |
{
"abstract": " Maximizing the sum of two generalized Rayleigh quotients (SRQ) can be\nreformulated as a one-dimensional optimization problem, where the function\nvalue evaluations are reduced to solving semi-definite programming (SDP)\nsubproblems. In this paper, we first use the dual SDP subproblem to construct\nan explicit overestimation and then propose a branch-and-bound algorithm to\nglobally solve (SRQ). Numerical results demonstrate that it is even more\nefficient than the recent SDP-based heuristic algorithm.\n",
"title": "An efficient global optimization algorithm for maximizing the sum of two generalized Rayleigh quotients"
}
| null | null | null | null | true | null |
6555
| null |
Default
| null | null |
null |
{
"abstract": " Dirichlet processes (DP) are widely applied in Bayesian nonparametric\nmodeling. However, in their basic form they do not directly integrate\ndependency information among data arising from space and time. In this paper,\nwe propose location dependent Dirichlet processes (LDDP) which incorporate\nnonparametric Gaussian processes in the DP modeling framework to model such\ndependencies. We develop the LDDP in the context of mixture modeling, and\ndevelop a mean field variational inference algorithm for this mixture model.\nThe effectiveness of the proposed modeling framework is shown on an image\nsegmentation task.\n",
"title": "Location Dependent Dirichlet Processes"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6556
| null |
Validated
| null | null |
null |
{
"abstract": " Kontsevich designed a scheme to generate infinitesimal symmetries\n$\\dot{\\mathcal{P}} = \\mathcal{Q}(\\mathcal{P})$ of Poisson brackets\n$\\mathcal{P}$ on all affine manifolds $M^r$; every such deformation is encoded\nby oriented graphs on $n+2$ vertices and $2n$ edges. In particular, these\nsymmetries can be obtained by orienting sums of non-oriented graphs $\\gamma$ on\n$n$ vertices and $2n-2$ edges. The bi-vector flow $\\dot{\\mathcal{P}} =\n\\text{Or}(\\gamma)(\\mathcal{P})$ preserves the space of Poisson structures if\n$\\gamma$ is a cocycle with respect to the vertex-expanding differential in the\ngraph complex.\nA class of such cocycles $\\boldsymbol{\\gamma}_{2\\ell+1}$ is known to exist:\nmarked by $\\ell \\in \\mathbb{N}$, each of them contains a $(2\\ell+1)$-gon wheel\nwith a nonzero coefficient. At $\\ell=1$ the tetrahedron $\\boldsymbol{\\gamma}_3$\nitself is a cocycle; at $\\ell=2$ the Kontsevich--Willwacher pentagon-wheel\ncocycle $\\boldsymbol{\\gamma}_5$ consists of two graphs. We reconstruct the\nsymmetry $\\mathcal{Q}_5(\\mathcal{P}) =\n\\text{Or}(\\boldsymbol{\\gamma}_5)(\\mathcal{P})$ and verify that $\\mathcal{Q}_5$\nis a Poisson cocycle indeed:\n$[\\![\\mathcal{P},\\mathcal{Q}_5(\\mathcal{P})]\\!]\\doteq 0$ via\n$[\\![\\mathcal{P},\\mathcal{P}]\\!]=0$.\n",
"title": "Poisson brackets symmetry from the pentagon-wheel cocycle in the graph complex"
}
| null | null | null | null | true | null |
6557
| null |
Default
| null | null |
null |
{
"abstract": " Chern-Schwartz-MacPherson (CSM) classes generalize to singular and/or\nnoncompact varieties the classical total homology Chern class of the tangent\nbundle of a smooth compact complex manifold. The theory of CSM classes has been\nextended to the equivariant setting by Ohmoto. We prove that for an arbitrary\ncomplex projective manifold $X$, the homogenized, torus equivariant CSM class\nof a constructible function $\\varphi$ is the restriction of the characteristic\ncycle of $\\varphi$ via the zero section of the cotangent bundle of $X$. This\nextends to the equivariant setting results of Ginzburg and Sabbah. We\nspecialize $X$ to be a (generalized) flag manifold $G/B$. In this case CSM\nclasses are determined by a Demazure-Lusztig (DL) operator. We prove a `Hecke\northogonality' of CSM classes, determined by the DL operator and its\nPoincar{é} adjoint. We further use the theory of holonomic\n$\\mathcal{D}_X$-modules to show that the characteristic cycle of a Verma\nmodule, restricted to the zero section, gives the CSM class of the\ncorresponding Schubert cell. Since the Verma characteristic cycles naturally\nidentify with the Maulik and Okounkov's stable envelopes, we establish an\nequivalence between CSM classes and stable envelopes; this reproves results of\nRim{á}nyi and Varchenko. As an application, we obtain a Segre type formula\nfor CSM classes. In the non-equivariant case this formula is manifestly\npositive, showing that the expansion in the Schubert basis of the CSM class of\na Schubert cell is effective. This proves a previous conjecture by Aluffi and\nMihalcea, and it extends previous positivity results by J. Huh in the Grassmann\nmanifold case. Finally, we generalize all of this to partial flag manifolds\n$G/P$.\n",
"title": "Shadows of characteristic cycles, Verma modules, and positivity of Chern-Schwartz-MacPherson classes of Schubert cells"
}
| null | null | null | null | true | null |
6558
| null |
Default
| null | null |
null |
{
"abstract": " The current data explosion poses great challenges to the approximate\naggregation with an efficiency and accuracy. To address this problem, we\npropose a novel approach to calculate the aggregation answers with a high\naccuracy using only a small portion of the data. We introduce leverages to\nreflect individual differences in the samples from a statistical perspective.\nTwo kinds of estimators, the leverage-based estimator, and the sketch estimator\n(a \"rough picture\" of the aggregation answer), are in constraint relations and\niteratively improved according to the actual conditions until their difference\nis below a threshold. Due to the iteration mechanism and the leverages, our\napproach achieves a high accuracy. Moreover, some features, such as not\nrequiring recording the sampled data and easy to extend to various execution\nmodes (e.g., the online mode), make our approach well suited to deal with big\ndata. Experiments show that our approach has an extraordinary performance, and\nwhen compared with the uniform sampling, our approach can achieve high-quality\nanswers with only 1/3 of the same sample size.\n",
"title": "An Iterative Scheme for Leverage-based Approximate Aggregation"
}
| null | null | null | null | true | null |
6559
| null |
Default
| null | null |
null |
{
"abstract": " We consider a two player dynamic game played over $T \\leq \\infty$ periods. In\neach period each player chooses any probability distribution with support on\n$[0,1]$ with a given mean, where the mean is the realized value of the draw\nfrom the previous period. The player with the highest realization in the period\nachieves a payoff of $1$, and the other player, $0$; and each player seeks to\nmaximize the discounted sum of his per-period payoffs over the whole time\nhorizon. We solve for the unique subgame perfect equilibrium of this game, and\nestablish properties of the equilibrium strategies and payoffs in the limit.\nThe solution and comparative statics thereof provide insight about\nintertemporal choice with status concerns. In particular we find that patient\nplayers take fewer risks.\n",
"title": "A Game of Martingales"
}
| null | null | null | null | true | null |
6560
| null |
Default
| null | null |
null |
{
"abstract": " We apply the newly derived nonadiabatic golden-rule instanton theory to\nasymmetric models describing electron-transfer in solution. The models go\nbeyond the usual spin-boson description and have anharmonic free-energy\nsurfaces with different values for the reactant and product reorganization\nenergies. The instanton method gives an excellent description of the behaviour\nof the rate constant with respect to asymmetry for the whole range studied. We\nderive a general formula for an asymmetric version of Marcus theory based on\nthe classical limit of the instanton and find that this gives significant\ncorrections to the standard Marcus theory. A scheme is given to compute this\nrate based only on equilibrium simulations. We also compare the rate constants\nobtained by the instanton method with its classical limit to study the effect\nof tunnelling and other quantum nuclear effects. These quantum effects can\nincrease the rate constant by orders of magnitude.\n",
"title": "Effects of tunnelling and asymmetry for system-bath models of electron transfer"
}
| null | null |
[
"Physics"
] | null | true | null |
6561
| null |
Validated
| null | null |
null |
{
"abstract": " If robots are to become ubiquitous, they will need to be able to adapt to\ncomplex and dynamic environments. Robots that can adapt their bodies while\ndeployed might be flexible and robust enough to meet this challenge. Previous\nwork on dynamic robot morphology has focused on simulation, combining simple\nmodules, or switching between locomotion modes. Here, we present an alternative\napproach: a self-reconfigurable morphology that allows a single four-legged\nrobot to actively adapt the length of its legs to different environments. We\nreport the design of our robot, as well as the results of a study that verifies\nthe performance impact of self-reconfiguration. This study compares three\ndifferent control and morphology pairs under different levels of servo supply\nvoltage in the lab. We also performed preliminary tests in different\nuncontrolled outdoor environments to see if changes to the external environment\nsupports our findings in the lab. Our results show better performance with an\nadaptable body, lending evidence to the value of self-reconfiguration for\nquadruped robots.\n",
"title": "Self-Modifying Morphology Experiments with DyRET: Dynamic Robot for Embodied Testing"
}
| null | null | null | null | true | null |
6562
| null |
Default
| null | null |
null |
{
"abstract": " Multi-layer neural networks have lead to remarkable performance on many kinds\nof benchmark tasks in text, speech and image processing. Nonlinear parameter\nestimation in hierarchical models is known to be subject to overfitting. One\napproach to this overfitting and related problems (local minima, colinearity,\nfeature discovery etc.) is called dropout (Srivastava, et al 2014, Baldi et al\n2016). This method removes hidden units with a Bernoulli random variable with\nprobability $p$ over updates. In this paper we will show that Dropout is a\nspecial case of a more general model published originally in 1990 called the\nstochastic delta rule ( SDR, Hanson, 1990). SDR parameterizes each weight in\nthe network as a random variable with mean $\\mu_{w_{ij}}$ and standard\ndeviation $\\sigma_{w_{ij}}$. These random variables are sampled on each forward\nactivation, consequently creating an exponential number of potential networks\nwith shared weights. Both parameters are updated according to prediction error,\nthus implementing weight noise injections that reflect a local history of\nprediction error and efficient model averaging. SDR therefore implements a\nlocal gradient-dependent simulated annealing per weight converging to a bayes\noptimal network. Tests on standard benchmarks (CIFAR) using a modified version\nof DenseNet shows the SDR outperforms standard dropout in error by over 50% and\nin loss by over 50%. Furthermore, the SDR implementation converges on a\nsolution much faster, reaching a training error of 5 in just 15 epochs with\nDenseNet-40 compared to standard DenseNet-40's 94 epochs.\n",
"title": "Dropout is a special case of the stochastic delta rule: faster and more accurate deep learning"
}
| null | null | null | null | true | null |
6563
| null |
Default
| null | null |
null |
{
"abstract": " Kernel regression is a popular non-parametric fitting technique. It aims at\nlearning a function which estimates the targets for test inputs as precise as\npossible. Generally, the function value for a test input is estimated by a\nweighted average of the surrounding training examples. The weights are\ntypically computed by a distance-based kernel function and they strongly depend\non the distances between examples. In this paper, we first review the latest\ndevelopments of sparse metric learning and kernel regression. Then a novel\nkernel regression method involving sparse metric learning, which is called\nkernel regression with sparse metric learning (KR$\\_$SML), is proposed. The\nsparse kernel regression model is established by enforcing a mixed $(2,1)$-norm\nregularization over the metric matrix. It learns a Mahalanobis distance metric\nby a gradient descent procedure, which can simultaneously conduct\ndimensionality reduction and lead to good prediction results. Our work is the\nfirst to combine kernel regression with sparse metric learning. To verify the\neffectiveness of the proposed method, it is evaluated on 19 data sets for\nregression. Furthermore, the new method is also applied to solving practical\nproblems of forecasting short-term traffic flows. In the end, we compare the\nproposed method with other three related kernel regression methods on all test\ndata sets under two criterions. Experimental results show that the proposed\nmethod is much more competitive.\n",
"title": "Kernel Regression with Sparse Metric Learning"
}
| null | null | null | null | true | null |
6564
| null |
Default
| null | null |
null |
{
"abstract": " We study the classification problems over string data for hypotheses\nspecified by formulas of monadic second-order logic MSO. The goal is to design\nlearning algorithms that run in time polynomial in the size of the training\nset, independently of or at least sublinear in the size of the whole data set.\nWe prove negative as well as positive results. If the data set is an\nunprocessed string to which our algorithms have local access, then learning in\nsublinear time is impossible even for hypotheses definable in a small fragment\nof first-order logic. If we allow for a linear time pre-processing of the\nstring data to build an index data structure, then learning of MSO-definable\nhypotheses is possible in time polynomial in the size of the training set,\nindependently of the size of the whole data set.\n",
"title": "Learning MSO-definable hypotheses on string"
}
| null | null | null | null | true | null |
6565
| null |
Default
| null | null |
null |
{
"abstract": " The recent realization of two-dimensional (2D) synthetic spin-orbit (SO)\ncoupling opens a broad avenue to study novel topological states for ultracold\natoms. Here, we propose a new scheme to realize exotic chiral Fulde-Ferrell\nsuperfluid for ultracold fermions, with a generic theory being shown that the\ntopology of superfluid pairing phases can be determined from the normal states.\nThe main findings are two fold. First, a semimetal is driven by a new type of\n2D SO coupling whose realization is even simpler than the recent experiment,\nand can be tuned into massive Dirac fermion phases with or without inversion\nsymmetry. Without inversion symmetry the superfluid phase with nonzero pairing\nmomentum is favored under an attractive interaction. Furthermore, we show a\nfundamental theorem that the topology of a 2D chiral superfluid can be uniquely\ndetermined from the unpaired normal states, with which the topological chiral\nFulde-Ferrell superfluid with a broad topological region is predicted for the\npresent system. This generic theorem is also useful for condensed matter\nphysics and material science in search for new topological superconductors.\n",
"title": "From semimetal to chiral Fulde-Ferrell superfluids"
}
| null | null | null | null | true | null |
6566
| null |
Default
| null | null |
null |
{
"abstract": " Characterization of lung nodules as benign or malignant is one of the most\nimportant tasks in lung cancer diagnosis, staging and treatment planning. While\nthe variation in the appearance of the nodules remains large, there is a need\nfor a fast and robust computer aided system. In this work, we propose an\nend-to-end trainable multi-view deep Convolutional Neural Network (CNN) for\nnodule characterization. First, we use median intensity projection to obtain a\n2D patch corresponding to each dimension. The three images are then\nconcatenated to form a tensor, where the images serve as different channels of\nthe input image. In order to increase the number of training samples, we\nperform data augmentation by scaling, rotating and adding noise to the input\nimage. The trained network is used to extract features from the input image\nfollowed by a Gaussian Process (GP) regression to obtain the malignancy score.\nWe also empirically establish the significance of different high level nodule\nattributes such as calcification, sphericity and others for malignancy\ndetermination. These attributes are found to be complementary to the deep\nmulti-view CNN features and a significant improvement over other methods is\nobtained.\n",
"title": "TumorNet: Lung Nodule Characterization Using Multi-View Convolutional Neural Network with Gaussian Process"
}
| null | null | null | null | true | null |
6567
| null |
Default
| null | null |
null |
{
"abstract": " The motion of electrons in or near solids, liquids and gases can be tracked\nby forcing their ejection with attosecond x-ray pulses, derived from\nfemtosecond lasers. The momentum of these emitted electrons carries the imprint\nof the electronic state. Aberration corrected transmission electron microscopes\nhave observed individual atoms, and have sufficient energy sensitivity to\nquantify atom bonding and electronic configurations. Recent developments in\nultrafast electron microscopy and diffraction indicate that spatial and\ntemporal information can be collected simultaneously. In the present work, we\npush the capability of femtosecond transmission electron microscopy (fs-TEM)\ntowards that of the state of the art in ultrafast lasers and electron\nmicroscopes. This is anticipated to facilitate unprecedented elucidation of\nphysical, chemical and biological structural dynamics on electronic time and\nlength scales. The fs-TEM numerically studied employs a nanotip source,\nelectrostatic acceleration to 70 keV, magnetic lens beam transport and\nfocusing, a condenser-objective around the sample and a terahertz temporal\ncompressor, including space charge effects during propagation. With electron\nemission equivalent to a 20 fs laser pulse, we find a spatial resolution below\n10 nm and a temporal resolution of below 10 fs will be feasible for pulses\ncomprised of on average 20 electrons. The influence of a transverse electric\nfield at the sample is modelled, indicating that a field of 1 V/$\\mu$m can be\nresolved.\n",
"title": "An apparatus architecture for femtosecond transmission electron microscopy"
}
| null | null | null | null | true | null |
6568
| null |
Default
| null | null |
null |
{
"abstract": " We present a hardware mechanism called HourGlass to predictably share data in\na multi-core system where cores are explicitly designated as critical or\nnon-critical. HourGlass is a time-based cache coherence protocol for\ndual-critical multi-core systems that ensures worst-case latency (WCL) bounds\nfor memory requests originating from critical cores. Although HourGlass does\nnot provide either WCL or bandwidth guarantees for memory requests from\nnon-critical cores, it promotes the use of timers to improve its bandwidth\nutilization while still maintaining WCL bounds for critical cores. This\nencourages a trade-off between the WCL bounds for critical cores, and the\nimproved memory bandwidth for non-critical cores via timer configurations. We\nevaluate HourGlass using gem5, and with multithreaded benchmark suites\nincluding SPLASH-2, and synthetic workloads. Our results show that the WCL for\ncritical cores with HourGlass is always within the analytical WCL bounds, and\nprovides a tighter WCL bound on critical cores compared to the state-of-the-art\nreal-time cache coherence protocol. Further, we show that HourGlass enables a\ntrade-off between provable WCL bounds for critical cores, and improved\nbandwidth utilization for non-critical cores. The average-case performance of\nHourGlass is comparable to the state-of-the-art real-time cache coherence\nprotocol, and suffers a slowdown of 1.43x and 1.46x compared to the\nconventional MSI and MESI protocols.\n",
"title": "HourGlass: Predictable Time-based Cache Coherence Protocol for Dual-Critical Multi-Core Systems"
}
| null | null | null | null | true | null |
6569
| null |
Default
| null | null |
null |
{
"abstract": " We investigated frictional effects on the folding rates of a human telomerase\nhairpin (hTR HP) and H-type pseudoknot from the Beet Western Yellow Virus (BWYV\nPK) using simulations of the Three Interaction Site (TIS) model for RNA. The\nheat capacity from TIS model simulations, calculated using temperature replica\nexchange simulations, reproduces nearly quantitatively the available\nexperimental data for the hTR HP. The corresponding results for BWYV PK serve\nas predictions. We calculated the folding rates ($k_\\mathrm{F}$) from more than\n100 folding trajectories for each value of the solvent viscosity ($\\eta$) at a\nfixed salt concentration of 200 mM. By using the theoretical estimate\n($\\propto$$\\sqrt{N}$ where $N$ is the number of nucleotides) for folding free\nenergy barrier, $k_\\mathrm{F}$ data for both the RNAs are quantitatively fit\nusing one-dimensional Kramers' theory with two parameters specifying the\ncurvatures in the unfolded basin and the barrier top. In the high-friction\nregime ($\\eta\\gtrsim10^{-5}\\,\\textrm{Pa\\ensuremath{\\cdot}s}$), for both HP and\nPK, $k_\\mathrm{F}$s decrease as $1/\\eta$ whereas in the low friction regime,\n$k_\\mathrm{F}$ values increase as $\\eta$ increases, leading to a maximum\nfolding rate at a moderate viscosity\n($\\sim10^{-6}\\,\\textrm{Pa\\ensuremath{\\cdot}s}$), which is the Kramers turnover.\nFrom the fits, we find that the speed limit to RNA folding at water viscosity\nis between 1 and 4 $\\mathrm{\\mu s}$, which is in accord with our previous\ntheoretical prediction as well as results from several single molecule\nexperiments. Both the RNA constructs fold by parallel pathways. Surprisingly,\nwe find that the flux through the pathways could be altered by changing solvent\nviscosity, a prediction that is more easily testable in RNA than in proteins.\n",
"title": "Frictional Effects on RNA Folding: Speed Limit and Kramers Turnover"
}
| null | null | null | null | true | null |
6570
| null |
Default
| null | null |
null |
{
"abstract": " Object ranking or \"learning to rank\" is an important problem in the realm of\npreference learning. On the basis of training data in the form of a set of\nrankings of objects represented as feature vectors, the goal is to learn a\nranking function that predicts a linear order of any new set of objects. In\nthis paper, we propose a new approach to object ranking based on principles of\nanalogical reasoning. More specifically, our inference pattern is formalized in\nterms of so-called analogical proportions and can be summarized as follows:\nGiven objects $A,B,C,D$, if object $A$ is known to be preferred to $B$, and $C$\nrelates to $D$ as $A$ relates to $B$, then $C$ is (supposedly) preferred to\n$D$. Our method applies this pattern as a main building block and combines it\nwith ideas and techniques from instance-based learning and rank aggregation.\nBased on first experimental results for data sets from various domains (sports,\neducation, tourism, etc.), we conclude that our approach is highly competitive.\nIt appears to be specifically interesting in situations in which the objects\nare coming from different subdomains, and which hence require a kind of\nknowledge transfer.\n",
"title": "Learning to Rank based on Analogical Reasoning"
}
| null | null | null | null | true | null |
6571
| null |
Default
| null | null |
null |
{
"abstract": " Modeling spatial overdispersion requires point processes models with finite\ndimensional distributions that are overdisperse relative to the Poisson.\nFitting such models usually heavily relies on the properties of stationarity,\nergodicity, and orderliness. And, though processes based on negative binomial\nfinite dimensional distributions have been widely considered, they typically\nfail to simultaneously satisfy the three required properties for fitting.\nIndeed, it has been conjectured by Diggle and Milne that no negative binomial\nmodel can satisfy all three properties. In light of this, we change\nperspective, and construct a new process based on a different overdisperse\ncount model, the Generalized Waring Distribution. While comparably tractable\nand flexible to negative binomial processes, the Generalized Waring process is\nshown to possess all required properties, and additionally span the negative\nbinomial and Poisson processes as limiting cases. In this sense, the GW process\nprovides an approximate resolution to the conundrum highlighted by Diggle and\nMilne.\n",
"title": "Modeling Spatial Overdispersion with the Generalized Waring Process"
}
| null | null | null | null | true | null |
6572
| null |
Default
| null | null |
null |
{
"abstract": " We explore methods of producing adversarial examples on deep generative\nmodels such as the variational autoencoder (VAE) and the VAE-GAN. Deep learning\narchitectures are known to be vulnerable to adversarial examples, but previous\nwork has focused on the application of adversarial examples to classification\ntasks. Deep generative models have recently become popular due to their ability\nto model input data distributions and generate realistic examples from those\ndistributions. We present three classes of attacks on the VAE and VAE-GAN\narchitectures and demonstrate them against networks trained on MNIST, SVHN and\nCelebA. Our first attack leverages classification-based adversaries by\nattaching a classifier to the trained encoder of the target generative model,\nwhich can then be used to indirectly manipulate the latent representation. Our\nsecond attack directly uses the VAE loss function to generate a target\nreconstruction image from the adversarial example. Our third attack moves\nbeyond relying on classification or the standard loss for the gradient and\ndirectly optimizes against differences in source and target latent\nrepresentations. We also motivate why an attacker might be interested in\ndeploying such techniques against a target generative network.\n",
"title": "Adversarial examples for generative models"
}
| null | null | null | null | true | null |
6573
| null |
Default
| null | null |
null |
{
"abstract": " We study the estimation of the covariance matrix $\\Sigma$ of a\n$p$-dimensional normal random vector based on $n$ independent observations\ncorrupted by additive noise. Only a general nonparametric assumption is imposed\non the distribution of the noise without any sparsity constraint on its\ncovariance matrix. In this high-dimensional semiparametric deconvolution\nproblem, we propose spectral thresholding estimators that are adaptive to the\nsparsity of $\\Sigma$. We establish an oracle inequality for these estimators\nunder model miss-specification and derive non-asymptotic minimax convergence\nrates that are shown to be logarithmic in $n/\\log p$. We also discuss the\nestimation of low-rank matrices based on indirect observations as well as the\ngeneralization to elliptical distributions. The finite sample performance of\nthe threshold estimators is illustrated in a numerical example.\n",
"title": "Sparse covariance matrix estimation in high-dimensional deconvolution"
}
| null | null | null | null | true | null |
6574
| null |
Default
| null | null |
null |
{
"abstract": " In the multi-agent systems setting, this paper addresses continuous-time\ndistributed synchronization of columns of rotation matrices. More precisely, k\nspecific columns shall be synchronized and only the corresponding k columns of\nthe relative rotations between the agents are assumed to be available for the\ncontrol design. When one specific column is considered, the problem is\nequivalent to synchronization on the (d-1)-dimensional unit sphere and when all\nthe columns are considered, the problem is equivalent to synchronization on\nSO(d). We design dynamic control laws for these synchronization problems. The\ncontrol laws are based on the introduction of auxiliary variables in\ncombination with a QR-factorization approach. The benefit of this\nQR-factorization approach is that we can decouple the dynamics for the k\ncolumns from the remaining d-k ones. Under the control scheme, the closed loop\nsystem achieves almost global convergence to synchronization for quasi-strong\ninteraction graph topologies.\n",
"title": "Dynamic controllers for column synchronization of rotation matrices: a QR-factorization approach"
}
| null | null | null | null | true | null |
6575
| null |
Default
| null | null |
null |
{
"abstract": " Compact and portable in-situ NMR spectrometers which can be dipped in the\nliquid to be measured, and are easily maintained, with affordable coil\nconstructions and electronics, together with an apparatus to recover depleted\nmagnets are presented, that provide a new real-time processing method for NMR\nspectrum acquisition, that remains stable despite magnetic field fluctuations.\n",
"title": "The ELEGANT NMR Spectrometer"
}
| null | null | null | null | true | null |
6576
| null |
Default
| null | null |
null |
{
"abstract": " Using different methods for laying out a graph can lead to very different\nvisual appearances, with which the viewer perceives different information.\nSelecting a \"good\" layout method is thus important for visualizing a graph. The\nselection can be highly subjective and dependent on the given task. A common\napproach to selecting a good layout is to use aesthetic criteria and visual\ninspection. However, fully calculating various layouts and their associated\naesthetic metrics is computationally expensive. In this paper, we present a\nmachine learning approach to large graph visualization based on computing the\ntopological similarity of graphs using graph kernels. For a given graph, our\napproach can show what the graph would look like in different layouts and\nestimate their corresponding aesthetic metrics. An important contribution of\nour work is the development of a new framework to design graph kernels. Our\nexperimental study shows that our estimation calculation is considerably faster\nthan computing the actual layouts and their aesthetic metrics. Also, our graph\nkernels outperform the state-of-the-art ones in both time and accuracy. In\naddition, we conducted a user study to demonstrate that the topological\nsimilarity computed with our graph kernel matches perceptual similarity\nassessed by human users.\n",
"title": "What Would a Graph Look Like in This Layout? A Machine Learning Approach to Large Graph Visualization"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6577
| null |
Validated
| null | null |
null |
{
"abstract": " We introduce a two-step procedure, in the context of ultra-high dimensional\nadditive models, which aims to reduce the size of covariates vector and\ndistinguish linear and nonlinear effects among nonzero components. Our proposed\nscreening procedure, in the first step, is constructed based on the concept of\ncumulative distribution function and conditional expectation of response in the\nframework of marginal correlation. B-splines and empirical distribution\nfunctions are used to estimate the two above measures. The sure property of\nthis procedure is also established. In the second step, a double penalization\nbased procedure is applied to identify nonzero and linear components,\nsimultaneously. The performance of the designed method is examined by several\ntest functions to show its capabilities against competitor methods when errors\ndistribution are varied. Simulation studies imply that the proposed screening\nprocedure can be applied to the ultra-high dimensional data and well detect the\nin uential covariates. It is also demonstrate the superiority in comparison\nwith the existing methods. This method is also applied to identify most in\nuential genes for overexpression of a G protein-coupled receptor in mice.\n",
"title": "A sure independence screening procedure for ultra-high dimensional partially linear additive models"
}
| null | null | null | null | true | null |
6578
| null |
Default
| null | null |
null |
{
"abstract": " We introduce a new class of Monte Carlo based approximations of expectations\nof random variables such that their laws are only available via certain\ndiscretizations. Sampling from the discretized versions of these laws can\ntypically introduce a bias. In this paper, we show how to remove that bias, by\nintroducing a new version of multi-index Monte Carlo (MIMC) that has the added\nadvantage of reducing the computational effort, relative to i.i.d. sampling\nfrom the most precise discretization, for a given level of error. We cover\nextensions of results regarding variance and optimality criteria for the new\napproach. We apply the methodology to the problem of computing an unbiased\nmollified version of the solution of a partial differential equation with\nrandom coefficients. A second application concerns the Bayesian inference (the\nsmoothing problem) of an infinite dimensional signal modelled by the solution\nof a stochastic partial differential equation that is observed on a discrete\nspace grid and at discrete times. Both applications are complemented by\nnumerical simulations.\n",
"title": "Unbiased Multi-index Monte Carlo"
}
| null | null | null | null | true | null |
6579
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we study symmetry properties of the Hilbert transformation of\nseveral real variables in the Clifford algebra setting. In order to describe\nthe symmetry properties we introduce the group $r\\mathrm{Spin}(n)+\\mathbb{R}^n,\nr>0,$ which is essentially an extension of the ax+b group. The study concludes\nthat the Hilbert transformation has certain characteristic symmetry properties\nin terms of $r\\mathrm{Spin}(n)+\\mathbb{R}^n.$ In the present paper, for $n=2$\nand $3$ we obtain, explicitly, the induced spinor representations of the\n$r\\mathrm{Spin}(n)+\\mathbb{R}^n$ group. Then we decompose the natural\nrepresentation of $r\\mathrm{Spin}(n)+\\mathbb{R}^n$ into the direct sum of some\ntwo irreducible spinor representations, by which we characterize the Hilbert\ntransformation in $\\mathbb{R}^3$ and $\\mathbb{R}^2.$ Precisely, we show that a\nnontrivial skew operator is the Hilbert transformation if and only if it is\ninvariant under the action of the $r\\mathrm{Spin}(n)+\\mathbb{R}^n, n=2,3,$\ngroup.\n",
"title": "Hilbert Transformation and $r\\mathrm{Spin}(n)+\\mathbb{R}^n$ Group"
}
| null | null | null | null | true | null |
6580
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study the large-time behavior of solutions to a class of\npartially dissipative linear hyperbolic systems with applications in\nvelocity-jump processes in several dimensions. Given integers $n,d\\ge 1$, let\n$\\mathbf A:=(A^1,\\dots,A^d)\\in (\\mathbb R^{n\\times n})^d$ be a matrix-vector,\nwhere $A^j\\in\\mathbb R^{n\\times n}$, and let $B\\in \\mathbb R^{n\\times n}$ be\nnot required to be symmetric but have one single eigenvalue zero, we consider\nthe Cauchy problem for linear $n\\times n$ systems having the form\n\\begin{equation*}\n\\partial_{t}u+\\mathbf A\\cdot \\nabla_{\\mathbf x} u+Bu=0,\\qquad (\\mathbf\nx,t)\\in \\mathbb R^d\\times \\mathbb R_+. \\end{equation*} Under appropriate\nassumptions, we show that the solution $u$ is decomposed into\n$u=u^{(1)}+u^{(2)}$, where $u^{(1)}$ has the asymptotic profile which is the\nsolution, denoted by $U$, of a parabolic equation and $u^{(1)}-U$ decays at the\nrate $t^{-\\frac d2(\\frac 1q-\\frac 1p)-\\frac 12}$ as $t\\to +\\infty$ in any\n$L^p$-norm, and $u^{(2)}$ decays exponentially in $L^2$-norm, provided\n$u(\\cdot,0)\\in L^q(\\mathbb R^d)\\cap L^2(\\mathbb R^d)$ for $1\\le q\\le p\\le\n\\infty$. Moreover, $u^{(1)}-U$ decays at the optimal rate $t^{-\\frac d2(\\frac\n1q-\\frac 1p)-1}$ as $t\\to +\\infty$ if the system satisfies a symmetry property.\nThe main proofs are based on asymptotic expansions of the solution $u$ in the\nfrequency space and the Fourier analysis.\n",
"title": "Asymptotic limit and decay estimates for a class of dissipative linear hyperbolic systems in several dimensions"
}
| null | null | null | null | true | null |
6581
| null |
Default
| null | null |
null |
{
"abstract": " Let (G, \\mu) be a pair of a reductive group G over the p-adic integers and a\nminuscule cocharacter {\\mu} of G defined over an unramified extension. We\nintroduce and study \"(G, \\mu)-displays\" which generalize Zink's Witt vector\ndisplays. We use these to define certain Rapoport-Zink formal schemes purely\ngroup theoretically, i.e. without p-divisible groups.\n",
"title": "(G, μ)-displays and Rapoport-Zink spaces"
}
| null | null | null | null | true | null |
6582
| null |
Default
| null | null |
null |
{
"abstract": " Program synthesis is a class of regression problems where one seeks a\nsolution, in the form of a source-code program, mapping the inputs to their\ncorresponding outputs exactly. Due to its precise and combinatorial nature,\nprogram synthesis is commonly formulated as a constraint satisfaction problem,\nwhere input-output examples are encoded as constraints and solved with a\nconstraint solver. A key challenge of this formulation is scalability: while\nconstraint solvers work well with a few well-chosen examples, a large set of\nexamples can incur significant overhead in both time and memory. We describe a\nmethod to discover a subset of examples that is both small and representative:\nthe subset is constructed iteratively, using a neural network to predict the\nprobability of unchosen examples conditioned on the chosen examples in the\nsubset, and greedily adding the least probable example. We empirically evaluate\nthe representativeness of the subsets constructed by our method, and\ndemonstrate such subsets can significantly improve synthesis time and\nstability.\n",
"title": "Selecting Representative Examples for Program Synthesis"
}
| null | null | null | null | true | null |
6583
| null |
Default
| null | null |
null |
{
"abstract": " We describe new irreducible components of the Gieseker-Maruyama moduli scheme\n$\\mathcal{M}(3)$ of semistable rank 2 coherent sheaves with Chern classes\n$c_1=0,\\ c_2=3,\\ c_3=0$ on $\\mathbb{P}^3$, general points of which correspond\nto sheaves whose singular loci contain components of dimensions both 0 and 1.\nThese sheaves are produced by elementary transformations of stable reflexive\nrank 2 sheaves with $c_1=0,\\ c_2=2,\\ c_3=2$ or 4 along a disjoint union of a\nprojective line and a collection of points in $\\mathbb{P}^3$. The constructed\nfamilies of sheaves provide first examples of irreducible components of the\nGieseker-Maruyama moduli scheme such that their general sheaves have\nsingularities of mixed dimension.\n",
"title": "Semistable rank 2 sheaves with singularities of mixed dimension on $\\mathbb{P}^3$"
}
| null | null | null | null | true | null |
6584
| null |
Default
| null | null |
null |
{
"abstract": " Cell monolayers provide an interesting example of active matter, exhibiting a\nphase transition from a flowing to jammed state as they age. Here we report\nexperiments and numerical simulations illustrating how a jammed cellular layer\nrapidly reverts to a flowing state after a wound. Quantitative comparison\nbetween experiments and simulations shows that cells change their\nself-propulsion and alignement strength so that the system crosses a phase\ntransition line, which we characterize by finite-size scaling in an active\nparticle model. This wound-induced unjamming transition is found to occur\ngenerically in epithelial, endothelial and cancer cells.\n",
"title": "From jamming to collective cell migration through a boundary induced transition"
}
| null | null | null | null | true | null |
6585
| null |
Default
| null | null |
null |
{
"abstract": " Long Short-Term Memory networks trained with gradient descent and\nback-propagation have received great success in various applications. However,\npoint estimation of the weights of the networks is prone to over-fitting\nproblems and lacks important uncertainty information associated with the\nestimation. However, exact Bayesian neural network methods are intractable and\nnon-applicable for real-world applications. In this study, we propose an\napproximate estimation of the weights uncertainty using Ensemble Kalman Filter,\nwhich is easily scalable to a large number of weights. Furthermore, we optimize\nthe covariance of the noise distribution in the ensemble update step using\nmaximum likelihood estimation. To assess the proposed algorithm, we apply it to\noutlier detection in five real-world events retrieved from the Twitter\nplatform.\n",
"title": "An Approximate Bayesian Long Short-Term Memory Algorithm for Outlier Detection"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
6586
| null |
Validated
| null | null |
null |
{
"abstract": " The covering type of a space $X$ is defined as the minimal cardinality of a\ngood cover of a space that is homotopy equivalent to $X$. We derive estimates\nfor the covering type of $X$ in terms of other invariants of $X$, namely the\nranks of the homology groups, the multiplicative structure of the cohomology\nring and the Lusternik-Schnirelmann category of $X$. By relating the covering\ntype to the number of vertices of minimal triangulations of complexes and\ncombinatorial manifolds, we obtain, within a unified framework, several\nestimates which are either new or extensions of results that have been\npreviously obtained by ad hoc combinatorial arguments. Moreover, our methods\ngive results that are valid for entire homotopy classes of spaces.\n",
"title": "Estimates of covering type and the number of vertices of minimal triangulations"
}
| null | null | null | null | true | null |
6587
| null |
Default
| null | null |
null |
{
"abstract": " This paper deals with the study of principal Lyapunov exponents, principal\nFloquet subspaces, and exponential separation for positive random linear\ndynamical systems in ordered Banach spaces. The main contribution lies in the\nintroduction of a new type of exponential separation, called of type II,\nimportant for its application to nonautonomous random differential equations\nwith delay. Under weakened assumptions, the existence of an exponential\nseparation of type II in an abstract general setting is shown, and an\nillustration of its application to dynamical systems generated by scalar linear\nrandom delay differential equations with finite delay is given.\n",
"title": "Principal Floquet subspaces and exponential separations of type II with applications to random delay differential equations"
}
| null | null | null | null | true | null |
6588
| null |
Default
| null | null |
null |
{
"abstract": " It is generally difficult to predict the positions of mutations in genomic\nDNA at the nucleotide level. Retroviral DNA insertion is one mode of mutation,\nresulting in host infections that are difficult to treat. This mutation process\ninvolves the integration of retroviral DNA into the host-infected cellular\ngenomic DNA following the interaction between host DNA and a pre-integration\ncomplex consisting of retroviral DNA and integrase. Here, we report that\nretroviral insertion sites around a hotspot within the Zfp521 and N-myc genes\ncan be predicted by a periodic function that is deduced using the diffraction\nlattice model. In conclusion, the mutagenesis process is described by a\nbiophysical model for DNA-DNA interactions.\n",
"title": "DNA insertion mutations can be predicted by a periodic probability function"
}
| null | null | null | null | true | null |
6589
| null |
Default
| null | null |
null |
{
"abstract": " Machine learning has emerged as an invaluable tool in many research areas. In\nthe present work, we harness this power to predict highly accurate molecular\ninfrared spectra with unprecedented computational efficiency. To account for\nvibrational anharmonic and dynamical effects -- typically neglected by\nconventional quantum chemistry approaches -- we base our machine learning\nstrategy on ab initio molecular dynamics simulations. While these simulations\nare usually extremely time consuming even for small molecules, we overcome\nthese limitations by leveraging the power of a variety of machine learning\ntechniques, not only accelerating simulations by several orders of magnitude,\nbut also greatly extending the size of systems that can be treated. To this\nend, we develop a molecular dipole moment model based on environment dependent\nneural network charges and combine it with the neural network potentials of\nBehler and Parrinello. Contrary to the prevalent big data philosophy, we are\nable to obtain very accurate machine learning models for the prediction of\ninfrared spectra based on only a few hundreds of electronic structure reference\npoints. This is made possible through the introduction of a fully automated\nsampling scheme and the use of molecular forces during neural network potential\ntraining. We demonstrate the power of our machine learning approach by applying\nit to model the infrared spectra of a methanol molecule, n-alkanes containing\nup to 200 atoms and the protonated alanine tripeptide, which at the same time\nrepresents the first application of machine learning techniques to simulate the\ndynamics of a peptide. In all these case studies we find excellent agreement\nbetween the infrared spectra predicted via machine learning models and the\nrespective theoretical and experimental spectra.\n",
"title": "Machine Learning Molecular Dynamics for the Simulation of Infrared Spectra"
}
| null | null |
[
"Physics",
"Statistics"
] | null | true | null |
6590
| null |
Validated
| null | null |
null |
{
"abstract": " In this article, we develop methods for estimating a low rank tensor from\nnoisy observations on a subset of its entries to achieve both statistical and\ncomputational efficiencies. There have been a lot of recent interests in this\nproblem of noisy tensor completion. Much of the attention has been focused on\nthe fundamental computational challenges often associated with problems\ninvolving higher order tensors, yet very little is known about their\nstatistical performance. To fill in this void, in this article, we characterize\nthe fundamental statistical limits of noisy tensor completion by establishing\nminimax optimal rates of convergence for estimating a $k$th order low rank\ntensor under the general $\\ell_p$ ($1\\le p\\le 2$) norm which suggest\nsignificant room for improvement over the existing approaches. Furthermore, we\npropose a polynomial-time computable estimating procedure based upon power\niteration and a second-order spectral initialization that achieves the optimal\nrates of convergence. Our method is fairly easy to implement and numerical\nexperiments are presented to further demonstrate the practical merits of our\nestimator.\n",
"title": "Statistically Optimal and Computationally Efficient Low Rank Tensor Completion from Noisy Entries"
}
| null | null | null | null | true | null |
6591
| null |
Default
| null | null |
null |
{
"abstract": " In his seminal paper \"Formality conjecture\", M. Kontsevich introduced a graph\ncomplex $GC_{1ve}$ closely connected with the problem of constructing a\nformality quasi-isomorphism for Hochschild cochains. In this paper, we express\nthe cohomology of the full directed graph complex explicitly in terms of the\ncohomology of $GC_{1ve}$. Applications of our results include a recent work by\nthe first author which completely characterizes homotopy classes of formality\nquasi-isomorphisms for Hochschild cochains in the stable setting.\n",
"title": "The cohomology of the full directed graph complex"
}
| null | null | null | null | true | null |
6592
| null |
Default
| null | null |
null |
{
"abstract": " We propose new types of models of the appearance of small- and large scale\nstructures in media with memory, including a hyperbolic modification of the\nNavier-Stokes equations and a class of dynamical low-dimensional models with\nmemory effects. On the basis of computer modeling, the formation of the\nsmall-scale structures and collapses and the appearance of new chaotic\nsolutions are demonstrated. Possibilities of the application of some proposed\nmodels to the description of the burst-type processes and collapses o nthe Sun\nare discussed.\n",
"title": "Model equations and structures formation for the media with memory"
}
| null | null | null | null | true | null |
6593
| null |
Default
| null | null |
null |
{
"abstract": " In this work, we provide non-asymptotic, probabilistic guarantees for\nsuccessful sparse support recovery by the multiple sparse Bayesian learning\n(M-SBL) algorithm in the multiple measurement vector (MMV) framework. For joint\nsparse Gaussian sources, we show that M-SBL perfectly recovers their common\nnonzero support with arbitrarily high probability using only finitely many\nMMVs. In fact, the support error probability decays exponentially fast with the\nnumber of MMVs, with the decay rate depending on the restricted isometry\nproperty of the self Khatri-Rao product of the measurement matrix. Our analysis\ntheoretically confirms that M-SBL is capable of recovering supports of size as\nhigh as $\\mathcal{O}(m^2)$, where $m$ is the number of measurements per sparse\nvector. In contrast, popular MMV algorithms in compressed sensing such as\nsimultaneous orthogonal matching pursuit and row-LASSO can recover only\n$\\mathcal{O}(m)$ sized supports. In the special case of noiseless measurements,\nwe show that a single MMV suffices for perfect recovery of the $k$-sparse\nsupport in M-SBL, provided any $k + 1$ columns of the measurement matrix are\nlinearly independent. Unlike existing support recovery guarantees for M-SBL,\nour sufficient conditions are non-asymptotic in nature, and do not require the\northogonality of the nonzero rows of the joint sparse signals.\n",
"title": "On the Support Recovery of Jointly Sparse Gaussian Sources using Sparse Bayesian Learning"
}
| null | null | null | null | true | null |
6594
| null |
Default
| null | null |
null |
{
"abstract": " The transition from single-cell to multicellular behavior is important in\nearly development but rarely studied. The starvation-induced aggregation of the\nsocial amoeba Dictyostelium discoideum into a multicellular slug is known to\nresult from single-cell chemotaxis towards emitted pulses of cyclic adenosine\nmonophosphate (cAMP). However, how exactly do transient short-range chemical\ngradients lead to coherent collective movement at a macroscopic scale? Here, we\nuse a multiscale model verified by quantitative microscopy to describe\nwide-ranging behaviors from chemotaxis and excitability of individual cells to\naggregation of thousands of cells. To better understand the mechanism of\nlong-range cell-cell communication and hence aggregation, we analyze cell-cell\ncorrelations, showing evidence for self-organization at the onset of\naggregation (as opposed to following a leader cell). Surprisingly, cell\ncollectives, despite their finite size, show features of criticality known from\nphase transitions in physical systems. Application of external cAMP\nperturbations in our simulations near the sensitive critical point allows\nsteering cells into early aggregation and towards certain locations but not\nonce an aggregation center has been chosen.\n",
"title": "A Critical-like Collective State Leads to Long-range Cell Communication in Dictyostelium discoideum Aggregation"
}
| null | null |
[
"Quantitative Biology"
] | null | true | null |
6595
| null |
Validated
| null | null |
null |
{
"abstract": " We review aspects of twistor theory, its aims and achievements spanning\nthelast five decades. In the twistor approach, space--time is secondary with\nevents being derived objects that correspond to compact holomorphic curves in a\ncomplex three--fold -- the twistor space. After giving an elementary\nconstruction of this space we demonstrate how solutions to linear and nonlinear\nequations of mathematical physics: anti-self-duality (ASD) equations on\nYang--Mills, or conformal curvature can be encoded into twistor cohomology.\nThese twistor correspondences yield explicit examples of Yang--Mills, and\ngravitational instantons which we review. They also underlie the twistor\napproach to integrability: the solitonic systems arise as symmetry reductions\nof ASD Yang--Mills equations, and Einstein--Weyl dispersionless systems are\nreductions of ASD conformal equations.\nWe then review the holomorphic string theories in twistor and ambitwistor\nspaces, and explain how these theories give rise to remarkable new formulae for\nthe computation of quantum scattering amplitudes. Finally we discuss the\nNewtonian limit of twistor theory, and its possible role in Penrose's proposal\nfor a role of gravity in quantum collapse of a wave function.\n",
"title": "Twistor theory at fifty: from contour integrals to twistor strings"
}
| null | null | null | null | true | null |
6596
| null |
Default
| null | null |
null |
{
"abstract": " Probabilistic integration of a continuous dynamical system is a way of\nsystematically introducing model error, at scales no larger than errors\nintroduced by standard numerical discretisation, in order to enable thorough\nexploration of possible responses of the system to inputs. It is thus a\npotentially useful approach in a number of applications such as forward\nuncertainty quantification, inverse problems, and data assimilation. We extend\nthe convergence analysis of probabilistic integrators for deterministic\nordinary differential equations, as proposed by Conrad et al.\\ (\\textit{Stat.\\\nComput.}, 2016), to establish mean-square convergence in the uniform norm on\ndiscrete- or continuous-time solutions under relaxed regularity assumptions on\nthe driving vector fields and their induced flows. Specifically, we show that\nrandomised high-order integrators for globally Lipschitz flows and randomised\nEuler integrators for dissipative vector fields with polynomially-bounded local\nLipschitz constants all have the same mean-square convergence rate as their\ndeterministic counterparts, provided that the variance of the integration noise\nis not of higher order than the corresponding deterministic integrator. These\nand similar results are proven for probabilistic integrators where the random\nperturbations may be state-dependent, non-Gaussian, or non-centred random\nvariables.\n",
"title": "Strong convergence rates of probabilistic integrators for ordinary differential equations"
}
| null | null | null | null | true | null |
6597
| null |
Default
| null | null |
null |
{
"abstract": " We propose a generalisation for the notion of the centre of an algebra in the\nsetup of algebras graded by an arbitrary abelian group G.\nOur generalisation, which we call the G-centre, is designed to control the\nendomorphism category of the grading shift functors. We show that the G-centre\nis preserved by gradable derived equivalences given by tilting modules. We also\ndiscuss links with existing notions in superalgebra theory and apply our\nresults to derived equivalences of superalgebras.\n",
"title": "The G-centre and gradable derived equivalences"
}
| null | null | null | null | true | null |
6598
| null |
Default
| null | null |
null |
{
"abstract": " Random forests is a common non-parametric regression technique which performs\nwell for mixed-type data and irrelevant covariates, while being robust to\nmonotonic variable transformations. Existing random forest implementations\ntarget regression or classification. We introduce the RFCDE package for fitting\nrandom forest models optimized for nonparametric conditional density\nestimation, including joint densities for multiple responses. This enables\nanalysis of conditional probability distributions which is useful for\npropagating uncertainty and of joint distributions that describe relationships\nbetween multiple responses and covariates. RFCDE is released under the MIT\nopen-source license and can be accessed at this https URL .\nBoth R and Python versions, which call a common C++ library, are available.\n",
"title": "RFCDE: Random Forests for Conditional Density Estimation"
}
| null | null |
[
"Statistics"
] | null | true | null |
6599
| null |
Validated
| null | null |
null |
{
"abstract": " Witten's Gauged Linear $\\sigma$-Model (GLSM) unifies the Gromov-Witten theory\nand the Landau-Ginzburg theory, and provides a global perspective on mirror\nsymmetry. In this article, we summarize a mathematically rigorous construction\nof the GLSM in the geometric phase using methods from symplectic geometry.\n",
"title": "The symplectic approach of gauged linear $σ$-model"
}
| null | null |
[
"Mathematics"
] | null | true | null |
6600
| null |
Validated
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.