text
null | inputs
dict | prediction
null | prediction_agent
null | annotation
list | annotation_agent
null | multi_label
bool 1
class | explanation
null | id
stringlengths 1
5
| metadata
null | status
stringclasses 2
values | event_timestamp
null | metrics
null |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
null |
{
"abstract": " We study the correlators of irregular vertex operators in two-dimensional\nconformal field theory (CFT) in order to propose an exact analytic formula for\ncalculating numbers of partitions, that is:\n1) for given $N,k$, finding the total number $\\lambda(N|k)$ of length $k$\npartitions of $N$: $N=n_1+...+n_k;0<n_1\\leq{n_2}...\\leq{n_k}$.\n2) finding the total number $\\lambda(N)=\\sum_{k=1}^N\\lambda(N|k)$ of\npartitions of a natural number $N$\nWe propose an exact analytic expression for $\\lambda(N|k)$ by relating\ntwo-point short-distance correlation functions of irregular vertex operators in\n$c=1$ conformal field theory ( the form of the operators is established in this\npaper): with the first correlator counting the partitions in the upper\nhalf-plane and the second one obtained from the first correlator by conformal\ntransformations of the form $f(z)=h(z)e^{-{i\\over{z}}}$ where $h(z)$ is regular\nand non-vanishing at $z=0$. The final formula for $\\lambda(N|k)$ is given in\nterms of regularized ($\\epsilon$-ordered) finite series in the generalized\nhigher-derivative Schwarzians and incomplete Bell polynomials of the above\nconformal transformation at $z=i\\epsilon$ ($\\epsilon\\rightarrow{0}$)\n",
"title": "An Analytic Formula for Numbers of Restricted Partitions from Conformal Field Theory"
}
| null | null |
[
"Mathematics"
] | null | true | null |
19101
| null |
Validated
| null | null |
null |
{
"abstract": " We introduce a new variant of the game of Cops and Robbers played on graphs,\nwhere the robber is invisible unless outside the neighbor set of a cop. The\nhyperopic cop number is the corresponding analogue of the cop number, and we\ninvestigate bounds and other properties of this parameter. We characterize the\ncop-win graphs for this variant, along with graphs with the largest possible\nhyperopic cop number. We analyze the cases of graphs with diameter 2 or at\nleast 3, focusing on when the hyperopic cop number is at most one greater than\nthe cop number. We show that for planar graphs, as with the usual cop number,\nthe hyperopic cop number is at most 3. The hyperopic cop number is considered\nfor countable graphs, and it is shown that for connected chains of graphs, the\nhyperopic cop density can be any real number in $[0,1/2].$\n",
"title": "Hyperopic Cops and Robbers"
}
| null | null | null | null | true | null |
19102
| null |
Default
| null | null |
null |
{
"abstract": " We prove new pinching estimate for the inverse curvature flow of strictly\nconvex hypersurfaces in the space form $N$ of constant sectional curvature\n$K_N$ with speed given by $F^{-\\alpha}$, where $\\alpha\\in (0,1]$ for $K_N=0,-1$\nand $\\alpha=1$ for $K_N=1$, $F$ is a smooth, symmetric homogeneous of degree\none function which is inverse concave and has dual $F_*$ approaching zero on\nthe boundary of the positive cone $\\Gamma_+$. We show that the ratio of the\nlargest principal curvature to the smallest principal curvature of the flow\nhypersurface is controlled by its initial value. This can be used to prove the\nsmooth convergence of the flow.\n",
"title": "New pinching estimates for Inverse curvature flows in space forms"
}
| null | null |
[
"Mathematics"
] | null | true | null |
19103
| null |
Validated
| null | null |
null |
{
"abstract": " We present an easy-to-use, Python-based framework that allows a researcher to\nautomate their computational simulations. In particular the framework\nfacilitates assembling several long-running computations and producing various\nplots from the data produced by these computations. The framework makes it\npossible to reproduce every figure made for a publication with a single\ncommand. It also allows one to distribute the computations across a network of\ncomputers. The framework has been used to write research papers in numerical\ncomputing. This paper discusses the design of the framework, and the benefits\nof using it. The ideas presented are general and should help researchers\norganize their computations for better reproducibility.\n",
"title": "automan: a simple, Python-based, automation framework for numerical computing"
}
| null | null | null | null | true | null |
19104
| null |
Default
| null | null |
null |
{
"abstract": " An experiment conducted in the framework of the EUHIT project and designed to\ncharacterize large scale structures in an adverse pressure gradient boundary\nlayer flow is presented. Up to 16 sCMOS cameras were used in order to perform\nlarge scale turbulent boundary layer PIV measurements with a large field of\nview and appropriate spatial resolution. To access the span-wise / wall-normal\nsignature of the structures as well, stereoscopic PIV measurements in\nspan-wise/wall-normal planes were performed at specific stream-wise locations.\nTo complement these large field of view measurements, long-range micro-PIV,\ntime resolved near wall velocity profiles and film-based measurements were\nperformed in order to determine the wall-shear stress and its fluctuations at\nsome specific locations along the model.\n",
"title": "Extensive characterization of a high Reynolds number decelerating boundary layer using advanced optical metrology"
}
| null | null |
[
"Physics"
] | null | true | null |
19105
| null |
Validated
| null | null |
null |
{
"abstract": " Let $F_n$ denote the $n^{th}$ Fibonacci number relative to the initial\nconditions $F_0=0$ and $F_1=1$. Bach, Paudyal, and Remmel introduced Fibonacci\nanalogues of the Stirling numbers called Fibo-Stirling numbers of the first and\nsecond kind. These numbers serve as the connection coefficients between the\nFibo-falling factorial basis $\\{(x)_{\\downarrow_{F,n}}:n \\geq 0\\}$ and the\nFibo-rising factorial basis $\\{(x)_{\\uparrow_{F,n}}:n \\geq 0\\}$ which are\ndefined by $(x)_{\\downarrow_{F,0}} = (x)_{\\uparrow_{F,0}} = 1$ and for $k \\geq\n1$, $(x)_{\\downarrow_{F,k}} = x(x-F_1) \\cdots (x-F_{k-1})$ and\n$(x)_{\\uparrow_{F,k}} = x(x+F_1) \\cdots (x+F_{k-1})$. We gave a general rook\ntheory model which allowed us to give combinatorial interpretations of the\nFibo-Stirling numbers of the first and second kind.\nThere are two natural $q$-analogues of the falling and rising Fibo-factorial\nbasis. That is, let $[x]_q = \\frac{q^x-1}{q-1}$. Then we let\n$[x]_{\\downarrow_{q,F,0}} = \\overline{[x]}_{\\downarrow_{q,F,0}} =\n[x]_{\\uparrow_{q,F,0}} = \\overline{[x]}_{\\uparrow_{q,F,0}}=1$ and, for $k > 0$,\nwe let $[x]_{\\downarrow_{q,F,k}} = [x]_q [x-F_1]_q \\cdots [x-F_{k-1}]_q$,\n$\\overline{[x]}_{\\downarrow_{q,F,k}}= [x]_q ([x]_q-[F_1]_q) \\cdots\n([x]_q-[F_{k-1}]_q)$, $[x]_{\\uparrow_{q,F,k}}= [x]_q [x+F_1]_q \\cdots\n[x+F_{k-1}]_q$, and $\\overline{[x]}_{\\uparrow_{q,F,k}}= [x]_q ([x]_q+[F_1]_q)\n\\cdots ([x]_q+[F_{k-1}]_q)$.\nIn this paper, we show we can modify the rook theory model of Bach, Paudyal,\nand Remmel to give combinatorial interpretations for the two different types\n$q$-analogues of the Fibo-Stirling numbers which arise as the connection\ncoefficients between the two different $q$-analogues of the Fibonacci falling\nand rising factorial bases. \\end{abstract}\n",
"title": "Q-analogues of the Fibo-Stirling numbers"
}
| null | null |
[
"Mathematics"
] | null | true | null |
19106
| null |
Validated
| null | null |
null |
{
"abstract": " Based on their formation mechanisms, Dirac points in three-dimensional\nsystems can be classified as accidental or essential. The former can be further\ndistinguished into type-I and type-II, depending on whether the Dirac cone\nspectrum is completely tipped over along certain direction. Here, we predict\nthe coexistence of all three kinds of Dirac points in the low-energy band\nstructure of CaAgBi-family materials with a stuffed Wurtzite structure. Two\npairs of accidental Dirac points reside on the rotational axis, with one pair\nbeing type-I and the other pair type-II; while another essential Dirac point is\npinned at the high symmetry point on the Brillouin zone boundary. Due to broken\ninversion symmetry, the band degeneracy around accidental Dirac points is\ncompletely lifted except along the rotational axis, which may enable the\nsplitting of chiral carriers at a ballistic p-n junction with a double negative\nrefraction effect. We clarify their symmetry protections, and find both the\nDirac-cone and Fermi arc topological surface states.\n",
"title": "Hybrid Dirac Semimetal in CaAgBi Materials Family"
}
| null | null |
[
"Physics"
] | null | true | null |
19107
| null |
Validated
| null | null |
null |
{
"abstract": " We consider a class of fractional stochastic volatility models (including the\nso-called rough Bergomi model), where the volatility is a superlinear function\nof a fractional Gaussian process. We show that the stock price is a true\nmartingale if and only if the correlation $\\rho$ between the driving Brownian\nmotions of the stock and the volatility is nonpositive. We also show that for\neach $\\rho<0$ and $m> \\frac{1}{1-\\rho^2}$, the $m$-th moment of the stock\nprice is infinite at each positive time.\n",
"title": "On the martingale property in the rough Bergomi model"
}
| null | null | null | null | true | null |
19108
| null |
Default
| null | null |
null |
{
"abstract": " A finite-dimensional algebra $A$ over an algebraically closed field $K$ is\ncalled periodic if it is periodic under the action of the syzygy operator in\nthe category of $A-A-$ bimodules. The periodic algebras are self-injective and\noccur naturally in the study of tame blocks of group algebras, actions of\nfinite groups on spheres, hypersurface singularities of finite Cohen-Macaulay\ntype, and Jacobian algebras of quivers with potentials. Recently, the tame\nperiodic algebras of polynomial growth have been classified and it is natural\nto attempt to classify all tame periodic algebras. We introduce the weighted\nsurface algebras of triangulated surfaces with arbitrarily oriented triangles\nand describe their basic properties. In particular, we prove that all these\nalgebras, except the singular tetrahedral algebras, are symmetric tame periodic\nalgebras of period $4$. Moreover, we describe the socle deformations of the\nweighted surface algebras and prove that all these algebras are symmetric tame\nperiodic algebras of period $4$. The main results of the paper form an\nimportant step towards a classification of all periodic symmetric tame algebras\nof non-polynomial growth, and lead to a complete description of all algebras of\ngeneralized quaternion type. Further, the orbit closures of the weighted\nsurface algebras (and their socle deformations) in the affine varieties of\nassociative $K$-algebra structures contain wide classes of tame symmetric\nalgebras related to algebras of dihedral and semidihedral types, which occur in\nthe study of blocks of group algebras with dihedral and semidihedral defect\ngroups.\n",
"title": "Weighted Surface Algebras"
}
| null | null | null | null | true | null |
19109
| null |
Default
| null | null |
null |
{
"abstract": " This paper studies the nonparametric modal regression problem systematically\nfrom a statistical learning view. Originally motivated by pursuing a\ntheoretical understanding of the maximum correntropy criterion based regression\n(MCCR), our study reveals that MCCR with a tending-to-zero scale parameter is\nessentially modal regression. We show that nonparametric modal regression\nproblem can be approached via the classical empirical risk minimization. Some\nefforts are then made to develop a framework for analyzing and implementing\nmodal regression. For instance, the modal regression function is described, the\nmodal regression risk is defined explicitly and its \\textit{Bayes} rule is\ncharacterized; for the sake of computational tractability, the surrogate modal\nregression risk, which is termed as the generalization risk in our study, is\nintroduced. On the theoretical side, the excess modal regression risk, the\nexcess generalization risk, the function estimation error, and the relations\namong the above three quantities are studied rigorously. It turns out that\nunder mild conditions, function estimation consistency and convergence may be\npursued in modal regression as in vanilla regression protocols, such as mean\nregression, median regression, and quantile regression. However, it outperforms\nthese regression models in terms of robustness as shown in our study from a\nre-descending M-estimation view. This coincides with and in return explains the\nmerits of MCCR on robustness. On the practical side, the implementation issues\nof modal regression including the computational algorithm and the tuning\nparameters selection are discussed. Numerical assessments on modal regression\nare also conducted to verify our findings empirically.\n",
"title": "A Statistical Learning Approach to Modal Regression"
}
| null | null | null | null | true | null |
19110
| null |
Default
| null | null |
null |
{
"abstract": " The ability to use a 2D map to navigate a complex 3D environment is quite\nremarkable, and even difficult for many humans. Localization and navigation is\nalso an important problem in domains such as robotics, and has recently become\na focus of the deep reinforcement learning community. In this paper we teach a\nreinforcement learning agent to read a map in order to find the shortest way\nout of a random maze it has never seen before. Our system combines several\nstate-of-the-art methods such as A3C and incorporates novel elements such as a\nrecurrent localization cell. Our agent learns to localize itself based on 3D\nfirst person images and an approximate orientation angle. The agent generalizes\nwell to bigger mazes, showing that it learned useful localization and\nnavigation capabilities.\n",
"title": "Teaching a Machine to Read Maps with Deep Reinforcement Learning"
}
| null | null | null | null | true | null |
19111
| null |
Default
| null | null |
null |
{
"abstract": " Average radio pulse profile of a pulsar B in a double pulsar system PSR\nJ0737-3039A/B exhibits an interesting behaviour. During the observation period\nbetween 2003 and 2009, the profile evolves from a single-peaked to a\ndouble-peaked form, following disappearance in 2008 indicating that the\ngeodetic precession of the pulsar is a possible origin of such behaviour. The\nknown pulsar beam models can be used to determine the geometry of PSR\nJ0737-3039B in the context of the precession. We study how the fan-beam\ngeometry performs in explaining the observed variations of the radio profile\nmorphology. It is shown that the fan beam can successfully reproduce the\nobserved evolution of the pulse width, and should be considered as a serious\nalternative for the conal-like models.\n",
"title": "The fan beam model for the pulse evolution of PSR J0737-3039B"
}
| null | null | null | null | true | null |
19112
| null |
Default
| null | null |
null |
{
"abstract": " Core-periphery networks are structures that present a set of central and\ndensely connected nodes, namely the core, and a set of non-central and sparsely\nconnected nodes, namely the periphery. The rich-club refers to a set in which\nthe highest degree nodes show a high density of connections. Thus, a network\nthat displays a rich-club can be interpreted as a core-periphery network in\nwhich the core is made up by a number of hubs. In this paper, we test the\nresilience of networks showing a progressively denser rich-club and we observe\nhow this structure is able to affect the network measures in terms of both\ncohesion and efficiency in information flow. Additionally, we consider the case\nin which, instead of making the core denser, we add links to the periphery.\nThese two procedures of core and periphery thickening delineate a decision\nprocess in the placement of new links and allow us to conduct a scenario\nanalysis that can be helpful in the comprehension and supervision of complex\nnetworks under the resilience perspective. The advantages of the two\nprocedures, as well as their implications, are discussed in relation to both\nnetwork effciency and node heterogeneity.\n",
"title": "Resilience of Core-Periphery Networks in the Case of Rich-Club"
}
| null | null | null | null | true | null |
19113
| null |
Default
| null | null |
null |
{
"abstract": " The last decade has seen a surge of interest in adaptive learning algorithms\nfor data stream classification, with applications ranging from predicting ozone\nlevel peaks, learning stock market indicators, to detecting computer security\nviolations. In addition, a number of methods have been developed to detect\nconcept drifts in these streams. Consider a scenario where we have a number of\nclassifiers with diverse learning styles and different drift detectors.\nIntuitively, the current 'best' (classifier, detector) pair is application\ndependent and may change as a result of the stream evolution. Our research\nbuilds on this observation. We introduce the $\\mbox{Tornado}$ framework that\nimplements a reservoir of diverse classifiers, together with a variety of drift\ndetection algorithms. In our framework, all (classifier, detector) pairs\nproceed, in parallel, to construct models against the evolving data streams. At\nany point in time, we select the pair which currently yields the best\nperformance. We further incorporate two novel stacking-based drift detection\nmethods, namely the $\\mbox{FHDDMS}$ and $\\mbox{FHDDMS}_{add}$ approaches. The\nexperimental evaluation confirms that the current 'best' (classifier, detector)\npair is not only heavily dependent on the characteristics of the stream, but\nalso that this selection evolves as the stream flows. Further, our\n$\\mbox{FHDDMS}$ variants detect concept drifts accurately in a timely fashion\nwhile outperforming the state-of-the-art.\n",
"title": "Reservoir of Diverse Adaptive Learners and Stacking Fast Hoeffding Drift Detection Methods for Evolving Data Streams"
}
| null | null | null | null | true | null |
19114
| null |
Default
| null | null |
null |
{
"abstract": " Regarding the analysis of Web communication, social and complex networks the\nfast finding of most influential nodes in a network graph constitutes an\nimportant research problem. We use two indices of the influence of those nodes,\nnamely, PageRank and a Max-linear model. We consider the PageRank %both as\n%Galton-Watson branching process and as an autoregressive process with a random\nnumber of random coefficients that depend on ranks of incoming nodes and their\nout-degrees and assume that the coefficients are independent and distributed\nwith regularly varying tail and with the same tail index. Then it is proved\nthat the tail index and the extremal index are the same for both PageRank and\nthe Max-linear model and the values of these indices are found. The\nachievements are based on the study of random sequences of a random length and\nthe comparison of the distribution of their maxima and linear combinations.\n",
"title": "Extremes in Random Graphs Models of Complex Networks"
}
| null | null | null | null | true | null |
19115
| null |
Default
| null | null |
null |
{
"abstract": " In recent work, redressed warped frames have been introduced for the analysis\nand synthesis of audio signals with non-uniform frequency and time resolutions.\nIn these frames, the allocation of frequency bands or time intervals of the\nelements of the representation can be uniquely described by means of a warping\nmap. Inverse warping applied after time-frequency sampling provides the key to\nreduce or eliminate dispersion of the warped frame elements in the conjugate\nvariable, making it possible, e.g., to construct frequency warped frames with\nsynchronous time alignment through frequency. The redressing procedure is\nhowever exact only when the analysis and synthesis windows have compact support\nin the domain where warping is applied. This implies that frequency warped\nframes cannot have compact support in the time domain. This property is\nundesirable when online computation is required. Approximations in which the\ntime support is finite are however possible, which lead to small reconstruction\nerrors. In this paper we study the approximation error for compactly supported\nfrequency warped analysis-synthesis elements, providing a few examples and case\nstudies.\n",
"title": "Estimates of the Reconstruction Error in Partially Redressed Warped Frames Expansions"
}
| null | null | null | null | true | null |
19116
| null |
Default
| null | null |
null |
{
"abstract": " For the multiterminal secret key agreement problem under a private source\nmodel, it is known that the maximum key rate, i.e., the secrecy capacity, can\nbe achieved through communication for omniscience, but the omniscience strategy\ncan be strictly suboptimal in terms of minimizing the public discussion rate.\nWhile a single-letter characterization is not known for the minimum discussion\nrate needed for achieving the secrecy capacity, we derive single-letter lower\nand upper bounds that yield some simple conditions for omniscience to be\ndiscussion-rate optimal. These conditions turn out to be enough to deduce the\noptimality of omniscience for a large class of sources including the\nhypergraphical sources. Through conjectures and examples, we explore other\nsource models to which our methods do not easily extend.\n",
"title": "On the Optimality of Secret Key Agreement via Omniscience"
}
| null | null | null | null | true | null |
19117
| null |
Default
| null | null |
null |
{
"abstract": " Constructing a smart wheelchair on a commercially available powered\nwheelchair (PWC) platform avoids a host of seating, mechanical design and\nreliability issues but requires methods of predicting and controlling the\nmotion of a device never intended for robotics. Analog joystick inputs are\nsubject to black-box transformations which may produce intuitive and adaptable\nmotion control for human operators, but complicate robotic control approaches;\nfurthermore, installation of standard axle mounted odometers on a commercial\nPWC is difficult. In this work, we present an integrated hardware and software\nsystem for predicting the motion of a commercial PWC platform that does not\nrequire any physical or electronic modification of the chair beyond plugging\ninto an industry standard auxiliary input port. This system uses an RGB-D\ncamera and an Arduino interface board to capture motion data, including visual\nodometry and joystick signals, via ROS communication. Future motion is\npredicted using an autoregressive sparse Gaussian process model. We evaluate\nthe proposed system on real-world short-term path prediction experiments.\nExperimental results demonstrate the system's efficacy when compared to a\nbaseline neural network model.\n",
"title": "Learning Motion Predictors for Smart Wheelchair using Autoregressive Sparse Gaussian Process"
}
| null | null | null | null | true | null |
19118
| null |
Default
| null | null |
null |
{
"abstract": " Kernel methods have produced state-of-the-art results for a number of NLP\ntasks such as relation extraction, but suffer from poor scalability due to the\nhigh cost of computing kernel similarities between natural language structures.\nA recently proposed technique, kernelized locality-sensitive hashing (KLSH),\ncan significantly reduce the computational cost, but is only applicable to\nclassifiers operating on kNN graphs. Here we propose to use random subspaces of\nKLSH codes for efficiently constructing an explicit representation of NLP\nstructures suitable for general classification methods. Further, we propose an\napproach for optimizing the KLSH model for classification problems by\nmaximizing an approximation of mutual information between the KLSH codes\n(feature vectors) and the class labels. We evaluate the proposed approach on\nbiomedical relation extraction datasets, and observe significant and robust\nimprovements in accuracy w.r.t. state-of-the-art classifiers, along with\ndrastic (orders-of-magnitude) speedup compared to conventional kernel methods.\n",
"title": "Kernelized Hashcode Representations for Relation Extraction"
}
| null | null | null | null | true | null |
19119
| null |
Default
| null | null |
null |
{
"abstract": " When you see a person in a crowd, occluded by other persons, you miss visual\ninformation that can be used to recognize, re-identify or simply classify him\nor her. You can imagine its appearance given your experience, nothing more.\nSimilarly, AI solutions can try to hallucinate missing information with\nspecific deep learning architectures, suitably trained with people with and\nwithout occlusions. The goal of this work is to generate a complete image of a\nperson, given an occluded version in input, that should be a) without occlusion\nb) similar at pixel level to a completely visible people shape c) capable to\nconserve similar visual attributes (e.g. male/female) of the original one. For\nthe purpose, we propose a new approach by integrating the state-of-the-art of\nneural network architectures, namely U-nets and GANs, as well as discriminative\nattribute classification nets, with an architecture specifically designed to\nde-occlude people shapes. The network is trained to optimize a Loss function\nwhich could take into account the aforementioned objectives. As well we propose\ntwo datasets for testing our solution: the first one, occluded RAP, created\nautomatically by occluding real shapes of the RAP dataset (which collects also\nattributes of the people aspect); the second is a large synthetic dataset, AiC,\ngenerated in computer graphics with data extracted from the GTA video game,\nthat contains 3D data of occluded objects by construction. Results are\nimpressive and outperform any other previous proposal. This result could be an\ninitial step to many further researches to recognize people and their behavior\nin an open crowded world.\n",
"title": "Can Adversarial Networks Hallucinate Occluded People With a Plausible Aspect?"
}
| null | null | null | null | true | null |
19120
| null |
Default
| null | null |
null |
{
"abstract": " This article presents \"John\", an open-source software designed to help\ncollective free improvisation. It provides generated screen-scores running on\ndistributed, reactive web-browsers. The musicians can then concurrently edit\nthe scores in their own browser. John is used by ONE, a septet playing\nimprovised electro-acoustic music with digital musical instruments (DMI). One\nof the original features of John is that its design takes care of leaving the\nmusician's attention as free as possible. Firstly, a quick review of the\ncontext of screen-based scores will help situate this research in the history\nof contemporary music notation. Then I will trace back how improvisation\nsessions led to John's particular \"notational perspective\". A brief description\nof the software will precede a discussion about the various aspects guiding its\ndesign.\n",
"title": "John, the semi-conductor : a tool for comprovisation"
}
| null | null | null | null | true | null |
19121
| null |
Default
| null | null |
null |
{
"abstract": " The real energy spectrum from the $PT$-symmetric Hamiltonian $H = p^2 -\n(ix)^N$ with $x\\in\\mathbb{C}$ was examined within one pair of Stokes wedges in\n1998 by Bender and Boettcher. For this Hamiltonian we discuss the following\nthree questions. First, since their paper used a Runge-Kutta method to\nintegrate along a path at the center of the Stokes wedges to calculate\neigenvalues $E$ with high accuracy, we wonder if the same eigenvalues can be\nobtained if integrate along some other paths in different shapes. Second, what\nthe corresponding eigenfunctions look like? Should the eigenfunctions be\nindependent from the shapes of path or not? Third, since for large $N$ the\nHamiltonian contains many pairs of Stokes wedges symmetric with respect to the\nimaginary axis of $x$, thus multiple families of real energy spectrum can be\nobtained. What do they look like? Any relation among them?\n",
"title": "Eigenvalue and Eigenfunction for the $PT$-symmetric Potential $V = - (ix)^N$"
}
| null | null | null | null | true | null |
19122
| null |
Default
| null | null |
null |
{
"abstract": " Virtually all real-world networks are dynamical entities. In social networks,\nthe propensity of nodes to engage in social interactions (activity) and their\nchances to be selected by active nodes (attractiveness) are heterogeneously\ndistributed. Here, we present a time-varying network model where each node and\nthe dynamical formation of ties are characterised by these two features. We\nstudy how these properties affect random walk processes unfolding on the\nnetwork when the time scales describing the process and the network evolution\nare comparable. We derive analytical solutions for the stationary state and the\nmean first passage time of the process and we study cases informed by empirical\nobservations of social networks. Our work shows that previously disregarded\nproperties of real social systems such heterogeneous distributions of activity\nand attractiveness as well as the correlations between them, substantially\naffect the dynamical process unfolding on the network.\n",
"title": "Random walks on activity-driven networks with attractiveness"
}
| null | null | null | null | true | null |
19123
| null |
Default
| null | null |
null |
{
"abstract": " B cells develop high affinity receptors during the course of affinity\nmaturation, a cyclic process of mutation and selection. At the end of affinity\nmaturation, a number of cells sharing the same ancestor (i.e. in the same\n\"clonal family\") are released from the germinal center, their amino acid\nfrequency profile reflects the allowed and disallowed substitutions at each\nposition. These clonal-family-specific frequency profiles, called \"substitution\nprofiles\", are useful for studying the course of affinity maturation as well as\nfor antibody engineering purposes. However, most often only a single sequence\nis recovered from each clonal family in a sequencing experiment, making it\nimpossible to construct a clonal-family-specific substitution profile. Given\nthe public release of many high-quality large B cell receptor datasets, one may\nask whether it is possible to use such data in a prediction model for\nclonal-family-specific substitution profiles. In this paper, we present the\nmethod \"Substitution Profiles Using Related Families\" (SPURF), a penalized\ntensor regression framework that integrates information from a rich assemblage\nof datasets to predict the clonal-family-specific substitution profile for any\nsingle input sequence. Using this framework, we show that substitution profiles\nfrom similar clonal families can be leveraged together with simulated\nsubstitution profiles and germline gene sequence information to improve\nprediction. We fit this model on a large public dataset and validate the\nrobustness of our approach on an external dataset. Furthermore, we provide a\ncommand-line tool in an open-source software package\n(this https URL) implementing these ideas and providing easy\nprediction using our pre-fit models.\n",
"title": "Predicting B Cell Receptor Substitution Profiles Using Public Repertoire Data"
}
| null | null | null | null | true | null |
19124
| null |
Default
| null | null |
null |
{
"abstract": " Merging Mobile Edge Computing (MEC), which is an emerging paradigm to meet\nthe increasing computation demands from mobile devices, with the dense\ndeployment of Base Stations (BSs), is foreseen as a key step towards the next\ngeneration mobile networks. However, new challenges arise for designing energy\nefficient networks since radio access resources and computing resources of BSs\nhave to be jointly managed, and yet they are complexly coupled with traffic in\nboth spatial and temporal domains. In this paper, we address the challenge of\nincorporating MEC into dense cellular networks, and propose an efficient online\nalgorithm, called ENGINE (ENErgy constrained offloadINg and slEeping) which\nmakes joint computation offloading and BS sleeping decisions in order to\nmaximize the quality of service while keeping the energy consumption low. Our\nalgorithm leverages Lyapunov optimization technique, works online and achieves\na close-to-optimal performance without using future information. Our simulation\nresults show that our algorithm can effectively reduce energy consumption\nwithout sacrificing the user quality of service.\n",
"title": "Energy Efficient Mobile Edge Computing in Dense Cellular Networks"
}
| null | null | null | null | true | null |
19125
| null |
Default
| null | null |
null |
{
"abstract": " In the realm of Delone sets in locally compact, second countable, Hausdorff\ngroups, we develop a dynamical systems approach in order to study the\ncontinuity behavior of measured quantities arising from point sets. A special\nfocus is both on the autocorrelation, as well as on the density of states for\nrandom bounded operators. It is shown that for uniquely ergodic limit systems,\nthe latter measures behave continuously with respect to the Chabauty-Fell\nconvergence of hulls. In the special situation of Euclidean spaces, our results\ncomplement recent developments in describing spectra as topological limits: we\nshow that the measured quantities under consideration can be approximated via\nperiodic analogs.\n",
"title": "Delone dynamical systems and spectral convergence"
}
| null | null |
[
"Mathematics"
] | null | true | null |
19126
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper, we study biconservative surfaces with parallel normalized mean\ncurvature vector in $\\mathbb{E}^4$. We obtain complete local classification in\n$\\mathbb{E}^4$ for a biconservative PNMCV surface. We also give an example to\nshow the existence of PNMCV biconservative surfaces in $\\mathbb{E}^4$.\n",
"title": "On biconservative surfaces in Euclidean spaces"
}
| null | null |
[
"Mathematics"
] | null | true | null |
19127
| null |
Validated
| null | null |
null |
{
"abstract": " In an effort to increase the versatility of finite element codes, we explore\nthe possibility of automatically creating the Jacobian matrix necessary for the\ngradient-based solution of nonlinear systems of equations. Particularly, we aim\nto assess the feasibility of employing the automatic differentiation tool\nTAPENADE for this purpose on a large Fortran codebase that is the result of\nmany years of continuous development. As a starting point we will describe the\nspecial structure of finite element codes and the implications that this code\ndesign carries for an efficient calculation of the Jacobian matrix. We will\nalso propose a first approach towards improving the efficiency of such a\nmethod. Finally, we will present a functioning method for the automatic\nimplementation of the Jacobian calculation in a finite element software, but\nwill also point out important shortcomings that will have to be addressed in\nthe future.\n",
"title": "Automatic implementation of material laws: Jacobian calculation in a finite element code with TAPENADE"
}
| null | null | null | null | true | null |
19128
| null |
Default
| null | null |
null |
{
"abstract": " We investigate the asymptotic behavior of solutions of Hamilton-Jacobi\nequations with large drift term in an open subset of two-dimensional Euclidean\nspace. When the drift is given by $\\varepsilon^{-1} (H_{x_2}, -H_{x_1})$ of a\nHamiltonian $H$, with $\\varepsilon > 0$, we establish the convergence, as\n$\\varepsilon \\to 0+$, of solutions of the Hamilton-Jacobi equations and\nidentify the limit of the solutions as the solution of systems of ordinary\ndifferential equations on a graph. This result generalizes the previous one\nobtained by the author to the case where the Hamiltonian $H$ admits a\ndegenerate critical point and, as a consequence, the graph may have segments\nmore than four at a node.\n",
"title": "Asymptotic analysis for Hamilton-Jacobi equations with large drift term"
}
| null | null | null | null | true | null |
19129
| null |
Default
| null | null |
null |
{
"abstract": " For a point set of $n$ elements in the $d$-dimensional unit cube and a class\nof test sets we are interested in the largest volume of a test set which does\nnot contain any point. For all natural numbers $n$, $d$ and under the\nassumption of a $delta$-cover with cardinality $\\vert \\Gamma_\\delta \\vert$ we\nprove that there is a point set, such that the largest volume of such a test\nset without any point is bounded by $\\frac{\\log \\vert \\Gamma_\\delta \\vert}{n} +\n\\delta$. For axis-parallel boxes on the unit cube this leads to a volume of at\nmost $\\frac{4d}{n}\\log(\\frac{9n}{d})$ and on the torus to $\\frac{4d}{n}\\log\n(2n)$.\n",
"title": "An Upper Bound of the Minimal Dispersion via Delta Covers"
}
| null | null | null | null | true | null |
19130
| null |
Default
| null | null |
null |
{
"abstract": " We have performed angle-resolved photoemission spectroscopy (ARPES) of LaSb\nand CeSb, a candidate of topological insulator. Using soft-x-ray photons, we\nhave accurately determined the three-dimensional bulk band structure and\nrevealed that the band inversion at the Brillouin-zone corner - a prerequisite\nfor realizing topological-insulator phase - is absent in both LaSb and CeSb.\nMoreover, unlike the ARPES data obtained with soft-x-ray photons, those with\nvacuum ultraviolet (VUV) photons were found to suffer significant $k_z$\nbroadening. These results suggest that LaSb and CeSb are topologically trivial\nsemimetals, and unusual Dirac-cone-like states observed with VUV photons are\nnot of the topological origin.\n",
"title": "Three-dimensional band structure of LaSb and CeSb:Absence of band inversion"
}
| null | null | null | null | true | null |
19131
| null |
Default
| null | null |
null |
{
"abstract": " In the hierarchical formation model, galaxy clusters grow by accretion of\nsmaller groups or isolated galaxies. During the infall into the centre of a\ncluster, the properties of accreted galaxies change. In particular, both\nobservations and numerical simulations suggest that its dark matter halo is\nstripped by the tidal forces of the host.\nWe use galaxy-galaxy weak lensing to measure the average mass of dark matter\nhaloes of satellite galaxies as a function of projected distance to the centre\nof the host, for different stellar mass bins. Assuming that the stellar\ncomponent of the galaxy is less disrupted by tidal stripping, stellar mass can\nbe used as a proxy of the infall mass. We study the stellar to halo mass\nrelation of satellites as a function of the cluster-centric distance to measure\ntidal stripping.\nWe use the shear catalogues of the DES science verification archive, the\nCFHTLenS and the CFHT Stripe 82 (CS82) surveys, and we select satellites from\nthe redMaPPer catalogue of clusters. For galaxies located in the outskirts of\nclusters, we find a stellar to halo mass relation in good agreement with the\ntheoretical expectations from \\citet{moster2013} for central galaxies. In the\ncentre of the cluster, we find that this relation is shifted to smaller halo\nmass for a given stellar mass. We interpret this finding as further evidence\nfor tidal stripping of dark matter haloes in high density environments.\n",
"title": "Stellar-to-halo mass relation of cluster galaxies"
}
| null | null | null | null | true | null |
19132
| null |
Default
| null | null |
null |
{
"abstract": " Using a novel rewriting problem, we show that several natural decision\nproblems about finite automata are undecidable (i.e., recursively unsolvable).\nIn contrast, we also prove three related problems are decidable. We apply one\nresult to prove the undecidability of a related problem about k-automatic sets\nof rational numbers.\n",
"title": "Undecidability and Finite Automata"
}
| null | null | null | null | true | null |
19133
| null |
Default
| null | null |
null |
{
"abstract": " We construct two examples of invariant manifolds that despite being locally\nunstable at every point in the transverse direction are globally stable. Using\nnumerical simulations we show that these invariant manifolds temporarily repel\nnearby trajectories but act as global attractors. We formulate an explanation\nfor such global stability in terms of the `rate of rotation' of the stable and\nunstable eigenvectors spanning the normal subspace associated with each point\nof the invariant manifold. We discuss the role of this rate of rotation on the\ntransitions between the stable and unstable regimes.\n",
"title": "A globally stable attractor that is locally unstable everywhere"
}
| null | null | null | null | true | null |
19134
| null |
Default
| null | null |
null |
{
"abstract": " We determine barycentric coordinates of triangle centers in the elliptic\nplane. The main focus is put on centers that lie on lines whose euclidean limit\n(triangle excess --> 0) is the Euler line or the Brocard line. We also\ninvestigate curves which can serve in elliptic geometry as substitutes for the\neuclidean nine-point-circle, the first Lemoine circle or the apollonian\ncircles.\n",
"title": "On Centers and Central Lines of Triangles in the Elliptic Plane"
}
| null | null |
[
"Mathematics"
] | null | true | null |
19135
| null |
Validated
| null | null |
null |
{
"abstract": " During active learning, an effective stopping method allows users to limit\nthe number of annotations, which is cost effective. In this paper, a new\nstopping method called Predicted Change of F Measure will be introduced that\nattempts to provide the users an estimate of how much performance of the model\nis changing at each iteration. This stopping method can be applied with any\nbase learner. This method is useful for reducing the data annotation bottleneck\nencountered when building text classification systems.\n",
"title": "Stopping Active Learning based on Predicted Change of F Measure for Text Classification"
}
| null | null | null | null | true | null |
19136
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, a random clique network model to mimic the large clustering\ncoefficient and the modular structure that exist in many real complex networks,\nsuch as social networks, artificial networks, and protein interaction networks,\nis introduced by combining the random selection rule of the Erdös and Rényi\n(ER) model and the concept of cliques. We find that random clique networks\nhaving a small average degree differ from the ER network in that they have a\nlarge clustering coefficient and a power law clustering spectrum, while\nnetworks having a high average degree have similar properties as the ER model.\nIn addition, we find that the relation between the clustering coefficient and\nthe average degree shows a non-monotonic behavior and that the degree\ndistributions can be fit by multiple Poisson curves; we explain the origin of\nsuch novel behaviors and degree distributions.\n",
"title": "Statistical properties of random clique networks"
}
| null | null |
[
"Computer Science",
"Physics"
] | null | true | null |
19137
| null |
Validated
| null | null |
null |
{
"abstract": " The paucity of videos in current action classification datasets (UCF-101 and\nHMDB-51) has made it difficult to identify good video architectures, as most\nmethods obtain similar performance on existing small-scale benchmarks. This\npaper re-evaluates state-of-the-art architectures in light of the new Kinetics\nHuman Action Video dataset. Kinetics has two orders of magnitude more data,\nwith 400 human action classes and over 400 clips per class, and is collected\nfrom realistic, challenging YouTube videos. We provide an analysis on how\ncurrent architectures fare on the task of action classification on this dataset\nand how much performance improves on the smaller benchmark datasets after\npre-training on Kinetics.\nWe also introduce a new Two-Stream Inflated 3D ConvNet (I3D) that is based on\n2D ConvNet inflation: filters and pooling kernels of very deep image\nclassification ConvNets are expanded into 3D, making it possible to learn\nseamless spatio-temporal feature extractors from video while leveraging\nsuccessful ImageNet architecture designs and even their parameters. We show\nthat, after pre-training on Kinetics, I3D models considerably improve upon the\nstate-of-the-art in action classification, reaching 80.9% on HMDB-51 and 98.0%\non UCF-101.\n",
"title": "Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset"
}
| null | null | null | null | true | null |
19138
| null |
Default
| null | null |
null |
{
"abstract": " We introduce and study the game of \"Selfish Cops and Active Robber\" (SCAR)\nwhich can be seen as an multiplayer variant of the \"classic\" two-player Cops\nand Robbers (CR) game. In classic CR all cops are controlled by a single\nplayer, who has no preference over which cop captures the robber. In SCAR, on\nthe other hand, each of N-1 cops is controlled by a separate player, and a\nsingle robber is controlled by the N-th player; and the capturing cop player\nreceives a higher reward than the non-capturing ones. Consequently, SCAR is an\nN-player pursuit game on graphs, in which each cop player has an increased\nmotive to be the one who captures the robber. The focus of our study is the\nexistence and properties of SCAR Nash Equilibria (NE). In particular, we prove\nthat SCAR always has one NE in deterministic positional strategies and (for N\ngreater than two) another in deterministic nonpositional strategies.\nFurthermore, we study conditions which, at equilibrium, guarantee either\ncapture or escape of the robber and show that (because of the antagonism\nbetween the \"selfish\" cop players) the robber may, in certain SCAR\nconfigurations, be captured later than he would be in classic CR, or even not\ncaptured at all. Finally we define the selfish cop number of a graph and study\nits connection to the classic cop number.\n",
"title": "Selfish Cops and Active Robber: Multi-Player Pursuit Evasion on Graphs"
}
| null | null | null | null | true | null |
19139
| null |
Default
| null | null |
null |
{
"abstract": " The 2016 Snook Prize has been awarded to Diego Tapias, Alessandro Bravetti,\nand David Sanders for their paper -- Ergodicity of One-Dimensional Systems\nCoupled to the Logistic Thermostat. They introduced a relatively stiff\nhyperbolic tangent thermostat force and successfully tested its ability to\nreproduce Gibbs' canonical distribution for the harmonic oscillator, the\nquartic oscillator, and the Mexican Hat potentials. Their work constitutes an\neffective response to the 2016 Ian Snook Prize Award goal -- Finding ergodic\nalgorithms for Gibbs' canonical ensemble using a single thermostat variable. We\nconfirm their work here and highlight an interesting feature of the Mexican Hat\nproblem when it is solved with an adaptive integrator.\n",
"title": "Singly-Thermostated Ergodicity in Gibbs' Canonical Ensemble and the 2016 Ian Snook Prize Award"
}
| null | null | null | null | true | null |
19140
| null |
Default
| null | null |
null |
{
"abstract": " We provide a new computationally-efficient class of estimators for risk\nminimization. We show that these estimators are robust for general statistical\nmodels: in the classical Huber epsilon-contamination model and in heavy-tailed\nsettings. Our workhorse is a novel robust variant of gradient descent, and we\nprovide conditions under which our gradient descent variant provides accurate\nestimators in a general convex risk minimization problem. We provide specific\nconsequences of our theory for linear regression, logistic regression and for\nestimation of the canonical parameters in an exponential family. These results\nprovide some of the first computationally tractable and provably robust\nestimators for these canonical statistical models. Finally, we study the\nempirical performance of our proposed methods on synthetic and real datasets,\nand find that our methods convincingly outperform a variety of baselines.\n",
"title": "Robust Estimation via Robust Gradient Estimation"
}
| null | null | null | null | true | null |
19141
| null |
Default
| null | null |
null |
{
"abstract": " In biology, there are several questions that translate to combinatorial\nsearch. For example, vesicle traffic systems that move cargo within eukaryotic\ncells have been proposed to exhibit several graph properties such as three\nconnectivity. These properties are consequences of underlying biophysical\nconstraints. A natural question for biologists is: what are the possible\nnetworks for various combinations of those properties? In this paper, we\npresent novel SMT based encodings of the properties over vesicle traffic\nsystems and a tool that searches for the networks that satisfies the properties\nusing SMT solvers. In our experiments, we show that our tool can search for\nnetworks of sizes that are considered to be relevant by biologists.\n",
"title": "SMT Solving for Vesicle Traffic Systems in Cells"
}
| null | null | null | null | true | null |
19142
| null |
Default
| null | null |
null |
{
"abstract": " Multipartite viruses replicate through a puzzling evolutionary strategy.\nTheir genome is segmented into two or more parts, and encapsidated in separate\nparticles that appear to propagate independently. Completing the replication\ncycle, however, requires the full genome, so that a persistent infection of a\nhost requires the concurrent presence of several particles. This represents an\napparent evolutionary drawback of multipartitism, while its advantages remain\nunclear. A transition from monopartite to multipartite viral forms has been\ndescribed in vitro under conditions of high multiplicity of infection,\nsuggesting that cooperation between defective mutants is a plausible\nevolutionary pathway towards multipartitism. However, it is unknown how the\nputative advantages that multipartitism might enjoy affect its epidemiology, or\nif an explicit advantage is needed to explain its ecological persistence. To\ndisentangle which mechanisms might contribute to the rise and fixation of\nmultipartitism, we here investigate the interaction between viral spreading\ndynamics and host population structure. We set up a compartmental model of the\nspread of a virus in its different forms and explore its epidemiology using\nboth analytical and numerical techniques. We uncover that the impact of host\ncontact structure on spreading dynamics entails a rich phenomenology of\necological relationships that includes cooperation, competition, and\ncommensality. Furthermore, we find out that multipartitism might rise to\nfixation even in the absence of explicit microscopic advantages. Multipartitism\nallows the virus to colonize environments that could not be invaded by the\nmonopartite form, while homogeneous contacts between hosts facilitate its\nspread. We conjecture that there might have been an increase in the diversity\nand prevalence of multipartite viral forms concomitantly with the expansion of\nagricultural practices.\n",
"title": "Endemicity and prevalence of multipartite viruses under heterogeneous between-host transmission"
}
| null | null | null | null | true | null |
19143
| null |
Default
| null | null |
null |
{
"abstract": " Collective cell migration is a highly regulated process involved in wound\nhealing, cancer metastasis and morphogenesis. Mechanical interactions among\ncells provide an important regulatory mechanism to coordinate such collective\nmotion. Using a Self-Propelled Voronoi (SPV) model that links cell mechanics to\ncell shape and cell motility, we formulate a generalized mechanical inference\nmethod to obtain the spatio-temporal distribution of cellular stresses from\nmeasured traction forces in motile tissues and show that such traction-based\nstresses match those calculated from instantaneous cell shapes. We additionally\nuse stress information to characterize the rheological properties of the\ntissue. We identify a motility-induced swim stress that adds to the interaction\nstress to determine the global contractility or extensibility of epithelia. We\nfurther show that the temporal correlation of the interaction shear stress\ndetermines an effective viscosity of the tissue that diverges at the\nliquid-solid transition, suggesting the possibility of extracting rheological\ninformation directly from traction data.\n",
"title": "Correlating Cell Shape and Cellular Stress in Motile Confluent Tissues"
}
| null | null | null | null | true | null |
19144
| null |
Default
| null | null |
null |
{
"abstract": " We consider the problem of learning function classes computed by neural\nnetworks with various activations (e.g. ReLU or Sigmoid), a task believed to be\ncomputationally intractable in the worst-case. A major open problem is to\nunderstand the minimal assumptions under which these classes admit provably\nefficient algorithms. In this work we show that a natural distributional\nassumption corresponding to {\\em eigenvalue decay} of the Gram matrix yields\npolynomial-time algorithms in the non-realizable setting for expressive classes\nof networks (e.g. feed-forward networks of ReLUs). We make no assumptions on\nthe structure of the network or the labels. Given sufficiently-strong\npolynomial eigenvalue decay, we obtain {\\em fully}-polynomial time algorithms\nin {\\em all} the relevant parameters with respect to square-loss. Milder decay\nassumptions also lead to improved algorithms. This is the first purely\ndistributional assumption that leads to polynomial-time algorithms for networks\nof ReLUs, even with one hidden layer. Further, unlike prior distributional\nassumptions (e.g., the marginal distribution is Gaussian), eigenvalue decay has\nbeen observed in practice on common data sets.\n",
"title": "Eigenvalue Decay Implies Polynomial-Time Learnability for Neural Networks"
}
| null | null |
[
"Computer Science"
] | null | true | null |
19145
| null |
Validated
| null | null |
null |
{
"abstract": " Steganography involves hiding a secret message or image inside another cover\nimage. Changes are made in the cover image without affecting visual quality of\nthe image. In contrast to cryptography, Steganography provides complete secrecy\nof the communication. Security of very sensitive data can be enhanced by\ncombining cryptography and steganography. A new technique that uses the concept\nof Steganography to obtain the position values from an image is suggested. This\npaper proposes a new method where no change is made to the cover image, only\nthe pixel position LSB (Least Significant Bit) values that match with the\nsecret message bit values are noted in a separate position file. At the sending\nend the position file along with the cover image is sent. At the receiving end\nthe position file is opened only with a secret key. The bit positions are taken\nfrom the position file and the LSB values from the positions are combined to\nget ASCII values and then form characters of the secret message\n",
"title": "A New Steganographic Technique Matching the Secret Message and Cover image Binary Value"
}
| null | null | null | null | true | null |
19146
| null |
Default
| null | null |
null |
{
"abstract": " Euler and Navier-Stokes have variant systems with dynamical invariance of\nhelicity and thus (weak) topological equivalence, allowing a strong `frozen-in'\n(to, or, dually, `Lie-carried' by the \\textit{virtual} velocity $V$)\nformulation of the vorticity with a flavor of `inverse Helmholtz theorem'. We\nremark on the non-ideal (statistical) topological fluid mechanics (TFM) for (1)\nthe Constantin-Iyer formulation of Navier-Stokes, (2) our own extension of the\nGallavotti-Cohen type dynamical ensembles of modified Navier-Stokes with\nenergy-helicity constraints and (3) the Galerkin truncated Euler, as the\ntypical case variants with dynamical time reversibility and helicity\ninvariance. Ideal TFM is thus bridged with non-ideal flows. An example virtual\n(Lie-)carrier of the vorticity in a Galerkin-truncated Euler system is\ncalculated to demonstrate the issue of determining $V$.\n",
"title": "On topological fluid mechanics of non-ideal systems and virtual frozen-in dynamics"
}
| null | null | null | null | true | null |
19147
| null |
Default
| null | null |
null |
{
"abstract": " We define a holographic dual to the Donaldson-Witten topological twist of\n$\\mathcal{N}=2$ gauge theories on a Riemannian four-manifold. This is described\nby a class of asymptotically locally hyperbolic solutions to $\\mathcal{N}=4$\ngauged supergravity in five dimensions, with the four-manifold as conformal\nboundary. Under AdS/CFT, minus the logarithm of the partition function of the\ngauge theory is identified with the holographically renormalized supergravity\naction. We show that the latter is independent of the metric on the boundary\nfour-manifold, as required for a topological theory. Supersymmetric solutions\nin the bulk satisfy first order differential equations for a twisted $Sp(1)$\nstructure, which extends the quaternionic Kahler structure that exists on any\nRiemannian four-manifold boundary. We comment on applications and extensions,\nincluding generalizations to other topological twists.\n",
"title": "Topological AdS/CFT"
}
| null | null | null | null | true | null |
19148
| null |
Default
| null | null |
null |
{
"abstract": " We systematically analyzed magnetodielectric resonances of Se colloids for\nthe first time to exploit the possibility for use as building blocks of\nall-dielectric optical metafluids. By taking synergistic advantages of Se\ncolloids, including (i) high-refractive-index at optical frequencies, (ii)\nunprecedented structural uniformity, and (iii) versatile access to copious\nquantities, the Kerker-type directional light scattering resulting from\nefficient coupling between strong electric and magnetic resonances were\nobserved directly from Se colloidal suspension. Thus, the use of Se colloid as\na generic magnetodielectric building block highlights an opportunity for the\nfluidic low-loss optical antenna, which can be processed via spin-coating and\npainting.\n",
"title": "Using highly uniform and smooth Selenium colloids as low-loss magnetodielectric building blocks of optical metafluids"
}
| null | null | null | null | true | null |
19149
| null |
Default
| null | null |
null |
{
"abstract": " We develop a commuting vector field method for a general class of radiating\nspacetimes. The metrics considered are certain long range perturbations of\nMinkowski space including those constructed from global stability problems in\ngeneral relativity. Our method provides sharp peeling estimates for solutions\nto both linear and nonlinear (null form) scalar fields.\n",
"title": "A Vector Field Method for Radiating Black Hole Spacetimes"
}
| null | null | null | null | true | null |
19150
| null |
Default
| null | null |
null |
{
"abstract": " We conducted a search for an exotic spin- and velocity-dependent interaction\nfor polarized electrons with an experimental approach based on a\nhigh-sensitivity spin-exchange relaxation-free (SERF) magnetometer, which\nserves as both a source of polarized electrons and a magnetic-field sensor. The\nexperiment aims to sensitively detect magnetic-fieldlike effects from the\nexotic interaction between the polarized electrons in a SERF vapor cell and\nunpolarized nucleons of a closely located solid-state mass. We report\nexperimental results on the interaction with 82 h of data averaging, which sets\nan experimental limit on the coupling strength around $10^{-19}$ for the axion\nmass $m_a \\lesssim 10^{-3}$ eV, within the important axion window.\n",
"title": "Experimental Constraint on an Exotic Spin- and Velocity-Dependent Interaction in the Sub-meV Range of Axion Mass with a Spin-Exchange Relaxation-Free Magnetometer"
}
| null | null | null | null | true | null |
19151
| null |
Default
| null | null |
null |
{
"abstract": " Deep learning models have lately shown great performance in various fields\nsuch as computer vision, speech recognition, speech translation, and natural\nlanguage processing. However, alongside their state-of-the-art performance, it\nis still generally unclear what is the source of their generalization ability.\nThus, an important question is what makes deep neural networks able to\ngeneralize well from the training set to new data. In this article, we provide\nan overview of the existing theory and bounds for the characterization of the\ngeneralization error of deep neural networks, combining both classical and more\nrecent theoretical and empirical results.\n",
"title": "Generalization Error in Deep Learning"
}
| null | null | null | null | true | null |
19152
| null |
Default
| null | null |
null |
{
"abstract": " Although block compressive sensing (BCS) makes it tractable to sense\nlarge-sized images and video, its recovery performance has yet to be\nsignificantly improved because its recovered images or video usually suffer\nfrom blurred edges, loss of details, and high-frequency oscillatory artifacts,\nespecially at a low subrate. This paper addresses these problems by designing a\nmodified total variation technique that employs multi-block gradient\nprocessing, a denoised Lagrangian multiplier, and patch-based sparse\nrepresentation. In the case of video, the proposed recovery method is able to\nexploit both spatial and temporal similarities. Simulation results confirm the\nimproved performance of the proposed method for compressive sensing of images\nand video in terms of both objective and subjective qualities.\n",
"title": "Block Compressive Sensing of Image and Video with Nonlocal Lagrangian Multiplier and Patch-based Sparse Representation"
}
| null | null | null | null | true | null |
19153
| null |
Default
| null | null |
null |
{
"abstract": " A consistent treatment of the coupling of surface energy and elasticity\nwithin the multi-phase- field framework is presented. The model accurately\nreproduces stress distribution in a number of analytically tractable, yet\nnon-trivial, cases including different types of spherical heterogeneities and a\nthin plate suspending in a gas environment. It is then used to study the stress\ndistribution inside elastic bodies with non-spherical geometries, such as a\nsolid ellipsoid and a sintered structure. In these latter cases, it is shown\nthat the interplay between deformation and spatially variable surface curvature\nleads to heterogeneous stress distribution across the specimen.\n",
"title": "A multi-phase-field method for surface tension-induced elasticity"
}
| null | null | null | null | true | null |
19154
| null |
Default
| null | null |
null |
{
"abstract": " We prove that recent breaking by Zahl of the $\\frac32$ barrier in Wolff's\nestimate on the Kakeya maximal operator in $\\mathbb R^4$ leads to improving the\n$\\frac{14}{5}$ threshold for the restriction problem for the paraboloid in\n$\\mathbb R^4$. One of the ingredients is a new trilinear estimate. The proofs\nare deliberately presented in a nontechnical and concise format, so as to make\nthe arguments more readable and focus attention on the key tools.\n",
"title": "On the restriction theorem for paraboloid in $\\mathbb R^4$"
}
| null | null | null | null | true | null |
19155
| null |
Default
| null | null |
null |
{
"abstract": " We establish existence of Stein kernels for probability measures on\n$\\mathbb{R}^d$ satisfying a Poincaré inequality, and obtain bounds on the\nStein discrepancy of such measures. Applications to quantitative central limit\ntheorems are discussed, including a new CLT in Wasserstein distance $W_2$ with\noptimal rate and dependence on the dimension. As a byproduct, we obtain a\nstability version of an estimate of the Poincaré constant of probability\nmeasures under a second moment constraint. The results extend more generally to\nthe setting of converse weighted Poincaré inequalities. The proof is based on\nsimple arguments of calculus of variations.\nFurther, we establish two general properties enjoyed by the Stein\ndiscrepancy, holding whenever a Stein kernel exists: Stein discrepancy is\nstrictly decreasing along the CLT, and it controls the skewness of a random\nvector.\n",
"title": "Existence of Stein Kernels under a Spectral Gap, and Discrepancy Bound"
}
| null | null | null | null | true | null |
19156
| null |
Default
| null | null |
null |
{
"abstract": " Synchronizations of processing elements (PEs) in massively parallel\nsimulations, which arise due to communication or load imbalances between PEs,\nsignificantly affect the scalability of scientific applications. We have\nrecently proposed a method based on finite-difference schemes to solve partial\ndifferential equations in an asynchronous fashion -- synchronization between\nPEs is relaxed at a mathematical level. While standard schemes can maintain\ntheir stability in the presence of asynchrony, their accuracy is drastically\naffected. In this work, we present a general methodology to derive\nasynchrony-tolerant (AT) finite difference schemes of arbitrary order of\naccuracy, which can maintain their accuracy when synchronizations are relaxed.\nWe show that there are several choices available in selecting a stencil to\nderive these schemes and discuss their effect on numerical and computational\nperformance. We provide a simple classification of schemes based on the stencil\nand derive schemes that are representative of different classes. Their\nnumerical error is rigorously analyzed within a statistical framework to obtain\nthe overall accuracy of the solution. Results from numerical experiments are\nused to validate the performance of the schemes.\n",
"title": "High-order asynchrony-tolerant finite difference schemes for partial differential equations"
}
| null | null | null | null | true | null |
19157
| null |
Default
| null | null |
null |
{
"abstract": " For future application of automated vehicles in public traffic, ensuring\nfunctional safety is essential. In this context, a hazard analysis and risk\nassessment is an important input for designing functionally vehicle automation\nsystems. In this contribution, we present a detailed hazard analysis and risk\nassessment (HARA) according to the ISO 26262 standard for a specific Level 4\napplication, namely an unmanned protective vehicle operated without human\nsupervision for motorway hard shoulder roadworks.\n",
"title": "Hazard Analysis and Risk Assessment for an Automated Unmanned Protective Vehicle"
}
| null | null |
[
"Computer Science"
] | null | true | null |
19158
| null |
Validated
| null | null |
null |
{
"abstract": " The Swift test was originally proposed as a formability test to reproduce the\nconditions observed in deep drawing operations. This test consists on forming a\ncylindrical cup from a circular blank, using a flat bottom cylindrical punch\nand has been extensively studied using both analytical and numerical methods.\nThis test can also be combined with the Demeri test, which consists in cutting\na ring from the wall of a cylindrical cup, in order to open it afterwards to\nmeasure the springback. This combination allows their use as benchmark test, in\norder to improve the knowledge concerning the numerical simulation models,\nthrough the comparison between experimental and numerical results. The focus of\nthis study is the experimental and numerical analyses of the Swift cup test,\nfollowed by the Demeri test, performed with an AA5754-O alloy at room\ntemperature. In this context, a detailed analysis of the punch force evolution,\nthe thickness evolution along the cup wall, the earing profile, the strain\npaths and their evolution and the ring opening is performed. The numerical\nsimulation is performed using the finite element code ABAQUS, with solid and\nsolid-shell elements, in order to compare the computational efficiency of these\ntype of elements. The results show that the solid-shell element is more\ncost-effective than the solid, presenting global accurate predictions, excepted\nfor the thinning zones. Both the von Mises and the Hill48 yield criteria\npredict the strain distributions in the final cup quite accurately. However,\nimproved knowledge concerning the stress states is still required, because the\nHill48 criterion showed difficulties in the correct prediction of the\nspringback, whatever the type of finite element adopted.\n",
"title": "Detailed experimental and numerical analysis of a cylindrical cup deep drawing: pros and cons of using solid-shell elements"
}
| null | null | null | null | true | null |
19159
| null |
Default
| null | null |
null |
{
"abstract": " Strong gravitational lensing by galaxy clusters is a fundamental tool to\nstudy dark matter and constrain the geometry of the Universe. Recently, the\nHubble Space Telescope Frontier Fields programme has allowed a significant\nimprovement of mass and magnification measurements but lensing models still\nhave a residual root mean square between 0.2 arcsec and few arcsec- onds, not\nyet completely understood. Systematic errors have to be better understood and\ntreated in order to use strong lensing clusters as reliable cosmological\nprobes. We have analysed two simulated Hubble-Frontier-Fields-like clusters\nfrom the Hubble Frontier Fields Comparison Challenge, Ares and Hera. We use\nseveral estimators (relative bias on magnification, den- sity profiles,\nellipticity and orientation) to quantify the goodness of our reconstructions by\ncomparing our multiple models, optimized with the parametric software LENSTOOL\n, with the input models. We have quantified the impact of systematic errors\narising, first, from the choice of different density profiles and\nconfigurations and, secondly, from the availability of con- straints\n(spectroscopic or photometric redshifts, redshift ranges of the background\nsources) in the parametric modelling of strong lensing galaxy clusters and\ntherefore on the retrieval of cosmological parameters. We find that\nsubstructures in the outskirts have a significant im- pact on the position of\nthe multiple images, yielding tighter cosmological contours. The need for\nwide-field imaging around massive clusters is thus reinforced. We show that\ncompetitive cosmological constraints can be obtained also with complex\nmultimodal clusters and that photometric redshifts improve the constraints on\ncosmological parameters when considering a narrow range of (spectroscopic)\nredshifts for the sources.\n",
"title": "Hubble Frontier Fields: systematic errors in strong lensing models of galaxy clusters - Implications for cosmography"
}
| null | null | null | null | true | null |
19160
| null |
Default
| null | null |
null |
{
"abstract": " E. Opdam introduced the tool of spectral transfer morphism (STM) of affine\nHecke algebras to study the formal degrees of unipotent discrete series\nrepresentations. He established a uniqueness property of STM for the affine\nHecke algebras associated of unipotent discrete series representations. Based\non this result, Opdam gave an explanation for Lusztig's arithmetic/geometric\ncorrespondence (in Lusztig's classification of unipotent representations of\n$p$-adic adjoint simple groups) in terms of harmonic analysis, and partitioned\nthe unipotent discrete series representations into $L$-packets based on the\nLusztig-Langlands parameters. The present paper provides some omitted details\nfor the argument of the uniqueness property of STM. In the last section, we\nprove that three finite morphisms of algebraic tori are spectral transfer\nmorphisms, and hence complete the proof of the uniqueness property.\n",
"title": "A Note on the Spectral Transfer Morphisms for Affine Hecke Algebras"
}
| null | null |
[
"Mathematics"
] | null | true | null |
19161
| null |
Validated
| null | null |
null |
{
"abstract": " We propose Batch-Expansion Training (BET), a framework for running a batch\noptimizer on a gradually expanding dataset. As opposed to stochastic\napproaches, batches do not need to be resampled i.i.d. at every iteration, thus\nmaking BET more resource efficient in a distributed setting, and when\ndisk-access is constrained. Moreover, BET can be easily paired with most batch\noptimizers, does not require any parameter-tuning, and compares favorably to\nexisting stochastic and batch methods. We show that when the batch size grows\nexponentially with the number of outer iterations, BET achieves optimal\n$O(1/\\epsilon)$ data-access convergence rate for strongly convex objectives.\nExperiments in parallel and distributed settings show that BET performs better\nthan standard batch and stochastic approaches.\n",
"title": "Batch-Expansion Training: An Efficient Optimization Framework"
}
| null | null | null | null | true | null |
19162
| null |
Default
| null | null |
null |
{
"abstract": " Coordinate descent methods usually minimize a cost function by updating a\nrandom decision variable (corresponding to one coordinate) at a time. Ideally,\nwe would update the decision variable that yields the largest decrease in the\ncost function. However, finding this coordinate would require checking all of\nthem, which would effectively negate the improvement in computational\ntractability that coordinate descent is intended to afford. To address this, we\npropose a new adaptive method for selecting a coordinate. First, we find a\nlower bound on the amount the cost function decreases when a coordinate is\nupdated. We then use a multi-armed bandit algorithm to learn which coordinates\nresult in the largest lower bound by interleaving this learning with\nconventional coordinate descent updates except that the coordinate is selected\nproportionately to the expected decrease. We show that our approach improves\nthe convergence of coordinate descent methods both theoretically and\nexperimentally.\n",
"title": "Coordinate Descent with Bandit Sampling"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
19163
| null |
Validated
| null | null |
null |
{
"abstract": " The set of 2-dimensional packing problems builds an important class of\noptimization problems and Strip Packing together with 2-dimensional Bin Packing\nand 2-dimensional Knapsack is one of the most famous of these problems. Given a\nset of rectangular axis parallel items and a strip with bounded width and\ninfinite height the objective is to find a packing of the items into the strip\nwhich minimizes the packing height. We speak of pseudo-polynomial Strip Packing\nif we consider algorithms with pseudo-polynomial running time with respect to\nthe width of the strip.\nIt is known that there is no pseudo-polynomial algorithm for Strip Packing\nwith a ratio better than $5/4$ unless $\\mathrm{P} = \\mathrm{NP}$. The best\nalgorithm so far has a ratio of $(4/3 + \\varepsilon)$. In this paper, we close\nthis gap between inapproximability result and best known algorithm by\npresenting an algorithm with approximation ratio $(5/4 + \\varepsilon)$ and thus\ncategorize the problem accurately. The algorithm uses a structural result which\nstates that each optimal solution can be transformed such that it has one of a\npolynomial number of different forms. The strength of this structural result is\nthat it applies to other problem settings as well for example to Strip Packing\nwith rotations (90 degrees) and Contiguous Moldable Task Scheduling. This fact\nenabled us to present algorithms with approximation ratio $(5/4 + \\varepsilon)$\nfor these problems as well.\n",
"title": "Closing the gap for pseudo-polynomial strip packing"
}
| null | null | null | null | true | null |
19164
| null |
Default
| null | null |
null |
{
"abstract": " A review of the replica symmetric solution of the classical and quantum,\ninfinite-range, Sherrington-Kirkpatrick spin glass is presented.\n",
"title": "Notes on the replica symmetric solution of the classical and quantum SK model, including the matrix of second derivatives and the spin glass susceptibility"
}
| null | null | null | null | true | null |
19165
| null |
Default
| null | null |
null |
{
"abstract": " To solve deep neural network (DNN)'s huge training dataset and its high\ncomputation issue, so-called teacher-student (T-S) DNN which transfers the\nknowledge of T-DNN to S-DNN has been proposed. However, the existing T-S-DNN\nhas limited range of use, and the knowledge of T-DNN is insufficiently\ntransferred to S-DNN. To improve the quality of the transferred knowledge from\nT-DNN, we propose a new knowledge distillation using singular value\ndecomposition (SVD). In addition, we define a knowledge transfer as a\nself-supervised task and suggest a way to continuously receive information from\nT-DNN. Simulation results show that a S-DNN with a computational cost of 1/5 of\nthe T-DNN can be up to 1.1\\% better than the T-DNN in terms of classification\naccuracy. Also assuming the same computational cost, our S-DNN outperforms the\nS-DNN driven by the state-of-the-art distillation with a performance advantage\nof 1.79\\%. code is available on this https URL\\_SVD.\n",
"title": "Self-supervised Knowledge Distillation Using Singular Value Decomposition"
}
| null | null | null | null | true | null |
19166
| null |
Default
| null | null |
null |
{
"abstract": " Hierarchy is an efficient way for a group to organize, but often goes along\nwith inequality that benefits leaders. To control despotic behaviour, followers\ncan assess leaders decisions by aggregating their own and their neighbours\nexperience, and in response challenge despotic leaders. But in hierarchical\nsocial networks, this interactional justice can be limited by (i) the high\ninfluence of a small clique who are treated better, and (ii) the low\nconnectedness of followers. Here we study how the connectedness of a social\nnetwork affects the co-evolution of despotism in leaders and tolerance to\ndespotism in followers. We simulate the evolution of a population of agents,\nwhere the influence of an agent is its number of social links. Whether a leader\nremains in power is controlled by the overall satisfaction of group members, as\ndetermined by their joint assessment of the leaders behaviour. We demonstrate\nthat centralization of a social network around a highly influential clique\ngreatly increases the level of despotism. This is because the clique is more\nsatisfied, and their higher influence spreads their positive opinion of the\nleader throughout the network. Finally, our results suggest that increasing the\nconnectedness of followers limits despotism while maintaining hierarchy.\n",
"title": "Can justice be fair when it is blind? How social network structures can promote or prevent the evolution of despotism"
}
| null | null |
[
"Quantitative Biology"
] | null | true | null |
19167
| null |
Validated
| null | null |
null |
{
"abstract": " Bioinformatics tools have been developed to interpret gene expression data at\nthe gene set level, and these gene set based analyses improve the biologists'\ncapability to discover functional relevance of their experiment design. While\nelucidating gene set individually, inter gene sets association is rarely taken\ninto consideration. Deep learning, an emerging machine learning technique in\ncomputational biology, can be used to generate an unbiased combination of gene\nset, and to determine the biological relevance and analysis consistency of\nthese combining gene sets by leveraging large genomic data sets. In this study,\nwe proposed a gene superset autoencoder (GSAE), a multi-layer autoencoder model\nwith the incorporation of a priori defined gene sets that retain the crucial\nbiological features in the latent layer. We introduced the concept of the gene\nsuperset, an unbiased combination of gene sets with weights trained by the\nautoencoder, where each node in the latent layer is a superset. Trained with\ngenomic data from TCGA and evaluated with their accompanying clinical\nparameters, we showed gene supersets' ability of discriminating tumor subtypes\nand their prognostic capability. We further demonstrated the biological\nrelevance of the top component gene sets in the significant supersets. Using\nautoencoder model and gene superset at its latent layer, we demonstrated that\ngene supersets retain sufficient biological information with respect to tumor\nsubtypes and clinical prognostic significance. Superset also provides high\nreproducibility on survival analysis and accurate prediction for cancer\nsubtypes.\n",
"title": "GSAE: an autoencoder with embedded gene-set nodes for genomics functional characterization"
}
| null | null |
[
"Statistics",
"Quantitative Biology"
] | null | true | null |
19168
| null |
Validated
| null | null |
null |
{
"abstract": " Given a nonconvex function that is an average of $n$ smooth functions, we\ndesign stochastic first-order methods to find its approximate stationary\npoints. The convergence of our new methods depends on the smallest (negative)\neigenvalue $-\\sigma$ of the Hessian, a parameter that describes how nonconvex\nthe function is.\nOur methods outperform known results for a range of parameter $\\sigma$, and\ncan be used to find approximate local minima. Our result implies an interesting\ndichotomy: there exists a threshold $\\sigma_0$ so that the currently fastest\nmethods for $\\sigma>\\sigma_0$ and for $\\sigma<\\sigma_0$ have different\nbehaviors: the former scales with $n^{2/3}$ and the latter scales with\n$n^{3/4}$.\n",
"title": "Natasha: Faster Non-Convex Stochastic Optimization Via Strongly Non-Convex Parameter"
}
| null | null |
[
"Computer Science",
"Mathematics",
"Statistics"
] | null | true | null |
19169
| null |
Validated
| null | null |
null |
{
"abstract": " We present a Bounded Model Checking technique for higher-order programs. The\nvehicle of our study is a higher-order calculus with general references. Our\ntechnique is a symbolic state syntactical translation based on SMT solvers,\nadapted to a setting where the values passed and stored during computation can\nbe functions of arbitrary order. We prove that our algorithm is sound, and\ndevise an optimisation based on points-to analysis to improve scalability. We\nmoreover provide a prototype implementation of the algorithm with experimental\nresults showcasing its performance.\n",
"title": "Higher-Order Bounded Model Checking"
}
| null | null | null | null | true | null |
19170
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we study the cubic fractional nonlinear Schrodinger equation\n(NLS) on the torus and on the real line. Combining the normal form and the\nrestricted norm methods we prove that the nonlinear part of the solution is\nsmoother than the initial data. Our method applies to both focusing and\ndefocusing nonlinearities. In the case of full dispersion (NLS) and on the\ntorus, the gain is a full derivative, while on the real line we get a\nderivative smoothing with an $\\epsilon$ loss. Our result lowers the regularity\nrequirement of a recent theorem of Kappeler et al. on the periodic defocusing\ncubic NLS, and extends it to the focusing case and to the real line. We also\nobtain estimates on the higher order Sobolev norms of the global smooth\nsolutions in the defocusing case.\n",
"title": "Smoothing for the fractional Schrodinger equation on the torus and the real line"
}
| null | null |
[
"Mathematics"
] | null | true | null |
19171
| null |
Validated
| null | null |
null |
{
"abstract": " Cryptovirological augmentations present an immediate, incomparable threat.\nOver the last decade, the substantial proliferation of crypto-ransomware has\nhad widespread consequences for consumers and organisations alike. Established\npreventive measures perform well, however, the problem has not ceased. Reverse\nengineering potentially malicious software is a cumbersome task due to platform\neccentricities and obfuscated transmutation mechanisms, hence requiring\nsmarter, more efficient detection strategies. The following manuscript presents\na novel approach for the classification of cryptographic primitives in compiled\nbinary executables using deep learning. The model blueprint, a DCNN, is\nfittingly configured to learn from variable-length control flow diagnostics\noutput from a dynamic trace. To rival the size and variability of contemporary\ndata compendiums, hence feeding the model cognition, a methodology for the\nprocedural generation of synthetic cryptographic binaries is defined, utilising\ncore primitives from OpenSSL with multivariate obfuscation, to draw a vastly\nscalable distribution. The library, CryptoKnight, rendered an algorithmic pool\nof AES, RC4, Blowfish, MD5 and RSA to synthesis combinable variants which are\nautomatically fed in its core model. Converging at 91% accuracy, CryptoKnight\nis successfully able to classify the sample algorithms with minimal loss.\n",
"title": "Deep Learning Based Cryptographic Primitive Classification"
}
| null | null |
[
"Computer Science"
] | null | true | null |
19172
| null |
Validated
| null | null |
null |
{
"abstract": " Various Alexandrov-Fenchel type inequalities have appeared and played\nimportant roles in convex geometry, matrix theory and complex algebraic\ngeometry. It has been noticed for some time that they share some striking\nanalogies and have intimate relationships. The purpose of this article is to\nshed new light on this by comparatively investigating them in several aspects.\n\\emph{The principal result} in this article is a complete solution to the\nequality characterization problem of various Alexandrov-Fenchel type\ninequalities for intersection numbers of nef and big classes on compact\nKähler manifolds, extending earlier results of Boucksom-Favre-Jonsson,\nFu-Xiao and Xiao-Lehmann. Our proof combines a result of Dinh-Nguyên on\nKähler geometry and an idea in convex geometry tracing back to Shephard. In\naddition to this central result, we also give a geometric proof of the complex\nversion of the Alexandrov-Fenchel type inequality for mixed discriminants and a\ndeterminantal type generalization of various Alexandrov-Fenchel type\ninequalities.\n",
"title": "The Alexandrov-Fenchel type inequalities, revisited"
}
| null | null | null | null | true | null |
19173
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we consider the role of nonmodal instabilities in the dynamics\nof atmospheric tornadoes. For this purpose we consider the Euler equation,\ncontinuity equation and the equation of state and linearise them. As an example\nwe study several different velocity profiles: the so-called Rankine vortex\nmodel; the Burgers-Rott vortex model; Sullivan and modified Sullivan vortex\nmodels. It has been shown that in the two dimensional Rankine vortex model no\ninstability appears in the inner region of a tornado. On the contrary, outside\nthis area the physical system undergoes strong exponential instability. We have\nfound that initially perturbed velocity components lead to amplified sound wave\nexcitations. The similar results have been shown in Burgers-Rott vortex model\nas well. As it was numerically estimated, in this case, the unstable wave\nincreases its energy by a factor of $400$ only in $\\sim 0.5$min. According to\nthe numerical study, in Sullivan and modified Sullivan models, the instability\ndoes not differ much by the growth. Despite the fact that in the inner area the\nexponential instability does not appear in a purely two dimensional case, we\nhave found that in the modified Sullivan vortex even a small contribution from\nvertical velocities can drive unstable nonmodal waves.\n",
"title": "Development of non-modal shear induced instabilities in atmospheric tornadoes"
}
| null | null | null | null | true | null |
19174
| null |
Default
| null | null |
null |
{
"abstract": " Face modeling has been paid much attention in the field of visual computing.\nThere exist many scenarios, including cartoon characters, avatars for social\nmedia, 3D face caricatures as well as face-related art and design, where\nlow-cost interactive face modeling is a popular approach especially among\namateur users. In this paper, we propose a deep learning based sketching system\nfor 3D face and caricature modeling. This system has a labor-efficient\nsketching interface, that allows the user to draw freehand imprecise yet\nexpressive 2D lines representing the contours of facial features. A novel CNN\nbased deep regression network is designed for inferring 3D face models from 2D\nsketches. Our network fuses both CNN and shape based features of the input\nsketch, and has two independent branches of fully connected layers generating\nindependent subsets of coefficients for a bilinear face representation. Our\nsystem also supports gesture based interactions for users to further manipulate\ninitial face models. Both user studies and numerical results indicate that our\nsketching system can help users create face models quickly and effectively. A\nsignificantly expanded face database with diverse identities, expressions and\nlevels of exaggeration is constructed to promote further research and\nevaluation of face modeling techniques.\n",
"title": "DeepSketch2Face: A Deep Learning Based Sketching System for 3D Face and Caricature Modeling"
}
| null | null | null | null | true | null |
19175
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we consider time-inhomogeneous branching processes and\ntime-inhomogeneous birth-and-death processes, in which the offspring\ndistribution and birth and death rates (respectively) vary in time. A classical\nresult of branching processes states that in the critical regime, a process\nconditioned on non-extinction and normalized will converge in distribution to a\nstandard exponential. In a paper of Jagers, time-inhomogeneous branching\nprocesses are shown to exhibit this convergence as well. In this paper, the\nhypotheses of Jagers' result are relaxed, further hypotheses are presented for\nconvergence in moments, and the result is extended to the continuous-time\nanalogue of time-inhomogeneous birth-and-death processes. In particular, the\nnew hypotheses suggest a simple characterization of the critical regime.\n",
"title": "Time-Inhomogeneous Branching Processes Conditioned on Non-Extinction"
}
| null | null | null | null | true | null |
19176
| null |
Default
| null | null |
null |
{
"abstract": " For a compact, connected metric graphs with a boundary that consists of $k$\nvertices, we prove that an arbitrary symmetric $k\\times k$ matrix with real\nentries can be realized as the Dirichlet-to-Neumann operator for the Laplacian\nplus a constant.\n",
"title": "The Dirichlet-to-Neumann operator for quantum graphs"
}
| null | null | null | null | true | null |
19177
| null |
Default
| null | null |
null |
{
"abstract": " The Monge-Kantorovich problem for the infinite Wasserstein distance presents\nseveral peculiarities. Among them the lack of convexity and then of a direct\nduality. We study in dimension 1 the dual problem introduced by Barron, Bocea\nand Jensen. We construct a couple of Kantorovich potentials which is \"as less\ntrivial as possible\". More precisely, we build a potential which is non\nconstant around any point that the plan which is locally optimal moves at\nmaximal distance. As an application, we show that the set of points which are\ndisplaced to maximal distance by a locally optimal transport plan is minimal.\n",
"title": "A study of the dual problem of the one-dimensional L-infinity optimal transport problem with applications"
}
| null | null | null | null | true | null |
19178
| null |
Default
| null | null |
null |
{
"abstract": " We use a non-perturbative renormalization group approach to develop a unified\npicture of the Bose polaron problem, where a mobile impurity is strongly\ninteracting with a surrounding Bose-Einstein condensate (BEC). A detailed\ntheoretical analysis of the phase diagram is presented and the\npolaron-to-molecule transition is discussed. For attractive polarons we argue\nthat a description in terms of an effective Fröhlich Hamiltonian with\nrenormalized parameters is possible. Its strong coupling regime is realized\nclose to a Feshbach resonance, where we predict a sharp increase of the\neffective mass. Already for weaker interactions, before the polaron mass\ndiverges, we predict a transition to a regime where states exist below the\npolaron energy and the attractive polaron is no longer the ground state. On the\nrepulsive side of the Feshbach resonance we recover the repulsive polaron,\nwhich has a finite lifetime because it can decay into low-lying molecular\nstates. We show for the entire range of couplings that the polaron energy has\nlogarithmic corrections in comparison with predictions by the mean-field\napproach. We demonstrate that they are a consequence of the polaronic mass\nrenormalization which is due to quantum fluctuations of correlated phonons in\nthe polaron cloud.\n",
"title": "Strong coupling Bose polarons in a BEC"
}
| null | null | null | null | true | null |
19179
| null |
Default
| null | null |
null |
{
"abstract": " Given a tournament T and a positive integer k, the C_3-Pakcing-T problem asks\nif there exists a least k (vertex-)disjoint directed 3-cycles in T. This is the\ndual problem in tournaments of the classical minimal feedback vertex set\nproblem. Surprisingly C_3-Pakcing-T did not receive a lot of attention in the\nliterature. We show that it does not admit a PTAS unless P=NP, even if we\nrestrict the considered instances to sparse tournaments, that is tournaments\nwith a feedback arc set (FAS) being a matching. Focusing on sparse tournaments\nwe provide a (1+6/(c-1)) approximation algorithm for sparse tournaments having\na linear representation where all the backward arcs have \"length\" at least c.\nConcerning kernelization, we show that C_3-Pakcing-T admits a kernel with O(m)\nvertices, where m is the size of a given feedback arc set. In particular, we\nderive a O(k) vertices kernel for C_3-Pakcing-T when restricted to sparse\ninstances. On the negative size, we show that C_3-Pakcing-T does not admit a\nkernel of (total bit) size O(k^{2-\\epsilon}) unless NP is a subset of coNP /\nPoly. The existence of a kernel in O(k) vertices for C_3-Pakcing-T remains an\nopen question.\n",
"title": "Triangle packing in (sparse) tournaments: approximation and kernelization"
}
| null | null | null | null | true | null |
19180
| null |
Default
| null | null |
null |
{
"abstract": " We prove Szegő-Widom asymptotics for the Chebyshev polynomials of a\ncompact subset of $\\mathbb{R}$ which is regular for potential theory and obeys\nthe Parreau-Widom and DCT conditions.\n",
"title": "Asymptotics of Chebyshev Polynomials, II. DCT Subsets of $\\mathbb{R}$"
}
| null | null | null | null | true | null |
19181
| null |
Default
| null | null |
null |
{
"abstract": " Timings of human activities are marked by circadian clocks which in turn are\nentrained to different environmental signals. In an urban environment the\npresence of artificial lighting and various social cues tend to disrupt the\nnatural entrainment with the sunlight. However, it is not completely understood\nto what extent this is the case. Here we exploit the large-scale data analysis\ntechniques to study the mobile phone calling activity of people in large cities\nto infer the dynamics of urban daily rhythms. From the calling patterns of\nabout 1,000,000 users spread over different cities but lying inside the same\ntime-zone, we show that the onset and termination of the calling activity\nsynchronizes with the east-west progression of the sun. We also find that the\nonset and termination of the calling activity of users follows a yearly\ndynamics, varying across seasons, and that its timings are entrained to solar\nmidnight. Furthermore, we show that the average mid-sleep time of people living\nin urban areas depends on the age and gender of each cohort as a result of\nbiological and social factors.\n",
"title": "Tracking Urban Human Activity from Mobile Phone Calling Patterns"
}
| null | null | null | null | true | null |
19182
| null |
Default
| null | null |
null |
{
"abstract": " In computational 3D geometric problems involving rotations, it is often that\npeople have to convert back and forth between a rotational matrix and a\nrotation described by an axis and a corresponding angle. For this purpose,\nRodrigues' rotation formula is a very popular expression to use because of its\nsimplicity and efficiency. Nevertheless, while converting a rotation matrix to\nan axis of rotation and the rotation angle, there exists ambiguity. Further\njudgement or even manual interference may be necessary in some situations. An\nextension of the Rodrigues' formula helps to find the sine and cosine values of\nthe rotation angle with respect to a given rotation axis is found and this\nsimple extension may help to accelerate many applications.\n",
"title": "Efficient conversion from rotating matrix to rotation axis and angle by extending Rodrigues' formula"
}
| null | null |
[
"Computer Science"
] | null | true | null |
19183
| null |
Validated
| null | null |
null |
{
"abstract": " Near infrared spectroscopy (NIRS) is an imaging-based diagnostic tool that\nprovides non-invasive and continuous evaluation of regional tissue oxygenation\nin real-time. In recent years, NIRS has show promise as a useful monitoring\ntechnology to help detect relative tissue ischemia that could lead to\nsignificant morbidity and mortality in preterm infants. However, some issues\ninherent in NIRS technology use on neonates, such as wide fluctuation in\nsignals, signal dropout and low limit of detection of the device, pose\nchallenges that may obscure reliable interpretation of the NIRS measurements\nusing current methods of analysis. In this paper, we propose new statistical\nmethods to analyse mesenteric rSO2 (regional oxygenation) produced by NIRS to\nevaluate oxygenation in intestinal tissues and investigate oxygenation response\nto red blood cell transfusion (RBC) in preterm infants. We present a mean area\nunder the curve (MAUC) measure and a slope measure to capture the mean rSO2\nlevel and temporal trajectory of rSO2, respectively. Estimation methods are\ndeveloped for these measures and nonparametric testing procedures are proposed\nto detect RBC-related changes in mesenteric oxygenation in preterm infants.\nThrough simulation studies, we show that the proposed methods demonstrate\nimproved accuracy in characterizing the mean level and changing pattern of\nmesenteric rSO2 and also increased statistical power in detecting RBC-related\nchanges, as compared with standard approaches. We apply our methods to a NIRS\nstudy in preterm infants receiving RBC transfusion from Emory Univerity to\nevaluate the pre- and post-transfusion mesenteric oxygenation in preterm\ninfants.\n",
"title": "Statistical methods for characterizing transfusion-related changes in regional oxygenation using Near-infrared spectroscopy (NIRS) in preterm infants"
}
| null | null | null | null | true | null |
19184
| null |
Default
| null | null |
null |
{
"abstract": " Generalization performance of classifiers in deep learning has recently\nbecome a subject of intense study. Deep models, typically over-parametrized,\ntend to fit the training data exactly. Despite this \"overfitting\", they perform\nwell on test data, a phenomenon not yet fully understood.\nThe first point of our paper is that strong performance of overfitted\nclassifiers is not a unique feature of deep learning. Using six real-world and\ntwo synthetic datasets, we establish experimentally that kernel machines\ntrained to have zero classification or near zero regression error perform very\nwell on test data, even when the labels are corrupted with a high level of\nnoise. We proceed to give a lower bound on the norm of zero loss solutions for\nsmooth kernels, showing that they increase nearly exponentially with data size.\nWe point out that this is difficult to reconcile with the existing\ngeneralization bounds. Moreover, none of the bounds produce non-trivial results\nfor interpolating solutions.\nSecond, we show experimentally that (non-smooth) Laplacian kernels easily fit\nrandom labels, a finding that parallels results for ReLU neural networks. In\ncontrast, fitting noisy data requires many more epochs for smooth Gaussian\nkernels. Similar performance of overfitted Laplacian and Gaussian classifiers\non test, suggests that generalization is tied to the properties of the kernel\nfunction rather than the optimization process.\nCertain key phenomena of deep learning are manifested similarly in kernel\nmethods in the modern \"overfitted\" regime. The combination of the experimental\nand theoretical results presented in this paper indicates a need for new\ntheoretical ideas for understanding properties of classical kernel methods. We\nargue that progress on understanding deep learning will be difficult until more\ntractable \"shallow\" kernel methods are better understood.\n",
"title": "To understand deep learning we need to understand kernel learning"
}
| null | null |
[
"Statistics"
] | null | true | null |
19185
| null |
Validated
| null | null |
null |
{
"abstract": " Measurements of element abundances in galaxies from astrophysical\nspectroscopy depend sensitively on the atomic data used. With the goal of\nmaking the latest atomic data accessible to the community, we present a\ncompilation of selected atomic data for resonant absorption lines at\nwavelengths longward of 911.753 {\\AA} (the \\ion{H}{1} Lyman limit), for key\nheavy elements (heavier than atomic number 5) of astrophysical interest. In\nparticular, we focus on the transitions of those ions that have been observed\nin the Milky Way interstellar medium (ISM), the circumgalactic medium (CGM) of\nthe Milky Way and/or other galaxies, and the intergalactic medium (IGM).\nWe provide wavelengths, oscillator strengths, associated accuracy grades, and\nreferences to the oscillator strength determinations. We also attempt to\ncompare and assess the recent oscillator strength determinations. For about\n22\\% of the lines that have updated oscillator strength values, the differences\nbetween the former values and the updated ones are $\\gtrsim$~0.1 dex.\nOur compilation will be a useful resource for absorption line studies of the\nISM, as well as studies of the CGM and IGM traced by sight lines to quasars and\ngamma-ray bursts. Studies (including those enabled by future generations of\nextremely large telescopes) of absorption by galaxies against the light of\nbackground galaxies will also benefit from our compilation.\n",
"title": "Atomic Data Revisions for Transitions Relevant to Observations of Interstellar, Circumgalactic, and Intergalactic Matter"
}
| null | null |
[
"Physics"
] | null | true | null |
19186
| null |
Validated
| null | null |
null |
{
"abstract": " With the aim of getting closer to the performance of the animal\nmuscleskeletal system, elastic elements are purposefully introduced in the\nmechanical structure of soft robots. Indeed, previous works have extensively\nshown that elasticity can endow robots with the ability of performing tasks\nwith increased efficiency, peak performances, and mechanical robustness.\nHowever, despite the many achievements, a general theory of efficient motions\nin soft robots is still lacking. Most of the literature focuses on specific\nexamples, or imposes a prescribed behavior through dynamic cancellations, thus\ndefeating the purpose of introducing elasticity in the first place. This paper\naims at making a step towards establishing such a general framework. To this\nend, we leverage on the theory of oscillations in nonlinear dynamical systems,\nand we take inspiration from state of the art theories about how the human\ncentral nervous system manages the muscleskeletal system. We propose to\ngenerate regular and efficient motions in soft robots by stabilizing\nsub-manifolds of the state space on which the system would naturally evolve. We\nselect these sub-manifolds as the nonlinear continuation of linear eigenspaces,\ncalled nonlinear normal modes. In such a way, efficient oscillatory behaviors\ncan be excited. We show the effectiveness of the methods in simulations on an\nelastic inverted pendulum, and experimentally on a segmented elastic leg.\n",
"title": "Using Nonlinear Normal Modes for Execution of Efficient Cyclic Motions in Soft Robots"
}
| null | null | null | null | true | null |
19187
| null |
Default
| null | null |
null |
{
"abstract": " Machine learning models benefit from large and diverse datasets. Using such\ndatasets, however, often requires trusting a centralized data aggregator. For\nsensitive applications like healthcare and finance this is undesirable as it\ncould compromise patient privacy or divulge trade secrets. Recent advances in\nsecure and privacy-preserving computation, including trusted hardware enclaves\nand differential privacy, offer a way for mutually distrusting parties to\nefficiently train a machine learning model without revealing the training data.\nIn this work, we introduce Myelin, a deep learning framework which combines\nthese privacy-preservation primitives, and use it to establish a baseline level\nof performance for fully private machine learning.\n",
"title": "Efficient Deep Learning on Multi-Source Private Data"
}
| null | null |
[
"Statistics"
] | null | true | null |
19188
| null |
Validated
| null | null |
null |
{
"abstract": " Starting from the integral representation of the three-dimensional Coulomb\ntransition matrix elaborated by us formerly with the use of specific symmetry\nof the interaction in a four-dimensional Euclidean space introduced by Fock,\nthe possibility of the analytical solving of the integral equation for the\npartial wave transition matrices at the excited bound state energy has been\nstudied. New analytical expressions for the partial s-, p- and d-wave Coulomb\nt-matrices for like-charged particles and the expression for the partial d-wave\nt-matrix for unlike-charged particles at the energy of the first excited bound\nstate have been derived.\n",
"title": "Analytical solution of the integral equation for partial wave Coulomb t-matrices at excited-state energy"
}
| null | null | null | null | true | null |
19189
| null |
Default
| null | null |
null |
{
"abstract": " We introduce the concept of $r$-equilateral $m$-gons. We prove the existence\nof $r$-equilateral $p$-gons in $\\mathbb R^d$ if $r<d$ and the existence of\nequilateral $p$-gons in the image of continuous injective maps $f:S^d\\to\n\\mathbb R^{d+1}$. Our ideas are based mainly in the paper of Y. Soibelman\n\\cite{soibelman}, in which the topological Borsuk number of $\\mathbb{R}^2$ is\ncalculated by means of topological methods and the paper of P. Blagojević and\nG. Ziegler \\cite{blagojevictetrahedra} where Fadell-Husseini index is used for\nsolving a problem related to the topological Borsuk problem for $\\mathbb{R}^3$.\n",
"title": "Equilateral $p$-gons in $\\mathbb R^d$ and deformed spheres and mod $p$ Fadell-Husseini index"
}
| null | null | null | null | true | null |
19190
| null |
Default
| null | null |
null |
{
"abstract": " According to Kearnes and Oman (2013), an ordered set $P$ is \\emph{Jónsson}\nif it is infinite and the cardinality of every proper initial segment of $P$ is\nstrictly less than the cardinaliy of $P$. We examine the structure of Jónsson\nposets.\n",
"title": "Jónsson posets"
}
| null | null |
[
"Mathematics"
] | null | true | null |
19191
| null |
Validated
| null | null |
null |
{
"abstract": " Drawing on some recent results that provide the formalism necessary to\ndefinite stationarity for infinite random graphs, this paper initiates the\nstudy of statistical and learning questions pertaining to these objects.\nSpecifically, a criterion for the existence of a consistent test for complex\nhypotheses is presented, generalizing the corresponding results on time series.\nAs an application, it is shown how one can test that a tree has the Markov\nproperty, or, more generally, to estimate its memory.\n",
"title": "Hypotheses testing on infinite random graphs"
}
| null | null | null | null | true | null |
19192
| null |
Default
| null | null |
null |
{
"abstract": " In the current article our primary objects of study are compact complex\nsubmanifolds of quotient manifolds of irreducible bounded symmetric domains by\ntorsion free discrete lattices of automorphisms. We are interested in the\ncharacterization of the totally geodesic submanifolds among compact splitting\ncomplex submanifolds, i.e. under the assumption that the tangent sequence\nsplits holomorphically over the submanifold.\n",
"title": "On compact splitting complex submanifolds of quotients of bounded symmetric domains"
}
| null | null | null | null | true | null |
19193
| null |
Default
| null | null |
null |
{
"abstract": " A third-order three-dimensional symmetric traceless tensor, called the\n\\emph{octupolar} tensor, has been introduced to study tetrahedratic nematic\nphases in liquid crystals. The octupolar \\emph{potential}, a scalar-valued\nfunction generated on the unit sphere by that tensor, should ideally have four\nmaxima capturing the most probable molecular orientations (on the vertices of a\ntetrahedron), but it was recently found to possess an equally generic variant\nwith \\emph{three} maxima instead of four. It was also shown that the\nirreducible admissible region for the octupolar tensor in a three-dimensional\nparameter space is bounded by a dome-shaped surface, beneath which is a\n\\emph{separatrix} surface connecting the two generic octupolar states. The\nlatter surface, which was obtained through numerical continuation, may be\nphysically interpreted as marking a possible \\emph{intra-octupolar} transition.\nIn this paper, by using the resultant theory of algebraic geometry and the\nE-characteristic polynomial of spectral theory of tensors, we give a\nclosed-form, algebraic expression for both the dome-shaped surface and the\nseparatrix surface. This turns the envisaged intra-octupolar transition into a\nquantitative, possibly observable prediction. Some other properties of\noctupolar tensors are also studied.\n",
"title": "Octupolar Tensors for Liquid Crystals"
}
| null | null | null | null | true | null |
19194
| null |
Default
| null | null |
null |
{
"abstract": " Strang splitting is a well established tool for the numerical integration of\nevolution equations. It allows the application of tailored integrators for\ndifferent parts of the vector field. However, it is also prone to order\nreduction in the case of non-trivial boundary conditions. This order reduction\ncan be remedied by correcting the boundary values of the intermediate splitting\nstep. In this paper, three different approaches for constructing such a\ncorrection in the case of inhomogeneous Dirichlet, Neumann, and mixed boundary\nconditions are presented. Numerical examples that illustrate the effectivity\nand benefits of these corrections are included.\n",
"title": "Efficient boundary corrected Strang splitting"
}
| null | null | null | null | true | null |
19195
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we introduce a new class of nonsmooth convex functions called\nSOS-convex semialgebraic functions extending the recently proposed notion of\nSOS-convex polynomials. This class of nonsmooth convex functions covers many\ncommon nonsmooth functions arising in the applications such as the Euclidean\nnorm, the maximum eigenvalue function and the least squares functions with\n$\\ell_1$-regularization or elastic net regularization used in statistics and\ncompressed sensing. We show that, under commonly used strict feasibility\nconditions, the optimal value and an optimal solution of SOS-convex\nsemi-algebraic programs can be found by solving a single semi-definite\nprogramming problem (SDP). We achieve the results by using tools from\nsemi-algebraic geometry, convex-concave minimax theorem and a recently\nestablished Jensen inequality type result for SOS-convex polynomials. As an\napplication, we outline how the derived results can be applied to show that\nrobust SOS-convex optimization problems under restricted spectrahedron data\nuncertainty enjoy exact SDP relaxations. This extends the existing exact SDP\nrelaxation result for restricted ellipsoidal data uncertainty and answers the\nopen questions left in [Optimization Letters 9, 1-18(2015)] on how to recover a\nrobust solution from the semi-definite programming relaxation in this broader\nsetting.\n",
"title": "SOS-convex Semi-algebraic Programs and its Applications to Robust Optimization: A Tractable Class of Nonsmooth Convex Optimization"
}
| null | null | null | null | true | null |
19196
| null |
Default
| null | null |
null |
{
"abstract": " We derive a Cramér-Rao lower bound for the variance of Floquet multiplier\nestimates that have been constructed from stable limit cycles perturbed by\nnoise. To do so, we consider perturbed periodic orbits in the plane. We use a\nperiodic autoregressive process to model the intersections of these orbits with\ncross sections, then passing to the limit of a continuum of sections to obtain\na bound that depends on the continuous flow restricted to the (nontrivial)\nFloquet mode. We compare our bound against the empirical variance of estimates\nconstructed using several cross sections. The section-based estimates are close\nto being optimal. We posit that the utility of our bound persists in higher\ndimensions when computed along Floquet modes for real and distinct multipliers.\nOur bound elucidates some of the empirical observations noted in the\nliterature; e.g., (a) it is the number of cycles (as opposed to the frequency\nof observations) that drives the variance of estimates to zero, and (b) the\nestimator variance has a positive lower bound as the noise amplitude tends to\nzero.\n",
"title": "An Uncertainty Principle for Estimates of Floquet Multipliers"
}
| null | null | null | null | true | null |
19197
| null |
Default
| null | null |
null |
{
"abstract": " Lack of moderation in online communities enables participants to incur in\npersonal aggression, harassment or cyberbullying, issues that have been\naccentuated by extremist radicalisation in the contemporary post-truth politics\nscenario. This kind of hostility is usually expressed by means of toxic\nlanguage, profanity or abusive statements. Recently Google has developed a\nmachine-learning-based toxicity model in an attempt to assess the hostility of\na comment; unfortunately, it has been suggested that said model can be deceived\nby adversarial attacks that manipulate the text sequence of the comment. In\nthis paper we firstly characterise such adversarial attacks as using\nobfuscation and polarity transformations. The former deceives by corrupting\ntoxic trigger content with typographic edits, whereas the latter deceives by\ngrammatical negation of the toxic content. Then, we propose a two--stage\napproach to counter--attack these anomalies, bulding upon a recently proposed\ntext deobfuscation method and the toxicity scoring model. Lastly, we conducted\nan experiment with approximately 24000 distorted comments, showing how in this\nway it is feasible to restore toxicity of the adversarial variants, while\nincurring roughly on a twofold increase in processing time. Even though novel\nadversary challenges would keep coming up derived from the versatile nature of\nwritten language, we anticipate that techniques combining machine learning and\ntext pattern recognition methods, each one targeting different layers of\nlinguistic features, would be needed to achieve robust detection of toxic\nlanguage, thus fostering aggression--free digital interaction.\n",
"title": "Shielding Google's language toxicity model against adversarial attacks"
}
| null | null | null | null | true | null |
19198
| null |
Default
| null | null |
null |
{
"abstract": " We develop numerical tools for Diagrammatic Monte-Carlo simulations of\nnon-Abelian lattice field theories in the t'Hooft large-N limit based on the\nweak-coupling expansion. First we note that the path integral measure of such\ntheories contributes a bare mass term in the effective action which is\nproportional to the bare coupling constant. This mass term renders the\nperturbative expansion infrared-finite and allows to study it directly in the\nlarge-N and infinite-volume limits using the Diagrammatic Monte-Carlo approach.\nOn the exactly solvable example of a large-N O(N) sigma model in D=2 dimensions\nwe show that this infrared-finite weak-coupling expansion contains, in addition\nto powers of bare coupling, also powers of its logarithm, reminiscent of\nre-summed perturbation theory in thermal field theory and resurgent\ntrans-series without exponential terms. We numerically demonstrate the\nconvergence of these double series to the manifestly non-perturbative dynamical\nmass gap. We then develop a Diagrammatic Monte-Carlo algorithm for sampling\nplanar diagrams in the large-N matrix field theory, and apply it to study this\ninfrared-finite weak-coupling expansion for large-N U(N)xU(N) nonlinear sigma\nmodel (principal chiral model) in D=2. We sample up to 12 leading orders of the\nweak-coupling expansion, which is the practical limit set by the increasingly\nstrong sign problem at high orders. Comparing Diagrammatic Monte-Carlo with\nconventional Monte-Carlo simulations extrapolated to infinite N, we find a good\nagreement for the energy density as well as for the critical temperature of the\n\"deconfinement\" transition. Finally, we comment on the applicability of our\napproach to planar QCD at zero and finite density.\n",
"title": "Diagrammatic Monte-Carlo for weak-coupling expansion of non-Abelian lattice field theories: large-N U(N)xU(N) principal chiral model"
}
| null | null | null | null | true | null |
19199
| null |
Default
| null | null |
null |
{
"abstract": " There is an increasing interest on accelerating neural networks for real-time\napplications. We study the student-teacher strategy, in which a small and fast\nstudent network is trained with the auxiliary information learned from a large\nand accurate teacher network. We propose to use conditional adversarial\nnetworks to learn the loss function to transfer knowledge from teacher to\nstudent. The proposed method is particularly effective for relatively small\nstudent networks. Moreover, experimental results show the effect of network\nsize when the modern networks are used as student. We empirically study the\ntrade-off between inference time and classification accuracy, and provide\nsuggestions on choosing a proper student network.\n",
"title": "Training Shallow and Thin Networks for Acceleration via Knowledge Distillation with Conditional Adversarial Networks"
}
| null | null |
[
"Computer Science"
] | null | true | null |
19200
| null |
Validated
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.