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": " In this work, we propose a simple but effective method to interpret black-box\nmachine learning models globally. That is, we use a compact binary tree, the\ninterpretation tree, to explicitly represent the most important decision rules\nthat are implicitly contained in the black-box machine learning models. This\ntree is learned from the contribution matrix which consists of the\ncontributions of input variables to predicted scores for each single\nprediction. To generate the interpretation tree, a unified process recursively\npartitions the input variable space by maximizing the difference in the average\ncontribution of the split variable between the divided spaces. We demonstrate\nthe effectiveness of our method in diagnosing machine learning models on\nmultiple tasks. Also, it is useful for new knowledge discovery as such insights\nare not easily identifiable when only looking at single predictions. In\ngeneral, our work makes it easier and more efficient for human beings to\nunderstand machine learning models.\n",
"title": "Global Model Interpretation via Recursive Partitioning"
}
| null | null | null | null | true | null |
15901
| null |
Default
| null | null |
null |
{
"abstract": " Adversarial learning of probabilistic models has recently emerged as a\npromising alternative to maximum likelihood. Implicit models such as generative\nadversarial networks (GAN) often generate better samples compared to explicit\nmodels trained by maximum likelihood. Yet, GANs sidestep the characterization\nof an explicit density which makes quantitative evaluations challenging. To\nbridge this gap, we propose Flow-GANs, a generative adversarial network for\nwhich we can perform exact likelihood evaluation, thus supporting both\nadversarial and maximum likelihood training. When trained adversarially,\nFlow-GANs generate high-quality samples but attain extremely poor\nlog-likelihood scores, inferior even to a mixture model memorizing the training\ndata; the opposite is true when trained by maximum likelihood. Results on MNIST\nand CIFAR-10 demonstrate that hybrid training can attain high held-out\nlikelihoods while retaining visual fidelity in the generated samples.\n",
"title": "Flow-GAN: Combining Maximum Likelihood and Adversarial Learning in Generative Models"
}
| null | null | null | null | true | null |
15902
| null |
Default
| null | null |
null |
{
"abstract": " Inference using deep neural networks is often outsourced to the cloud since\nit is a computationally demanding task. However, this raises a fundamental\nissue of trust. How can a client be sure that the cloud has performed inference\ncorrectly? A lazy cloud provider might use a simpler but less accurate model to\nreduce its own computational load, or worse, maliciously modify the inference\nresults sent to the client. We propose SafetyNets, a framework that enables an\nuntrusted server (the cloud) to provide a client with a short mathematical\nproof of the correctness of inference tasks that they perform on behalf of the\nclient. Specifically, SafetyNets develops and implements a specialized\ninteractive proof (IP) protocol for verifiable execution of a class of deep\nneural networks, i.e., those that can be represented as arithmetic circuits.\nOur empirical results on three- and four-layer deep neural networks demonstrate\nthe run-time costs of SafetyNets for both the client and server are low.\nSafetyNets detects any incorrect computations of the neural network by the\nuntrusted server with high probability, while achieving state-of-the-art\naccuracy on the MNIST digit recognition (99.4%) and TIMIT speech recognition\ntasks (75.22%).\n",
"title": "SafetyNets: Verifiable Execution of Deep Neural Networks on an Untrusted Cloud"
}
| null | null | null | null | true | null |
15903
| null |
Default
| null | null |
null |
{
"abstract": " We consider the factorization problem of matrix symbols relative to a closed\ncontour, i.e., a Riemann-Hilbert problem, where the symbol depends analytically\non parameters. We show how to define a function $\\tau$ which is locally\nanalytic on the space of deformations and that is expressed as a Fredholm\ndeterminant of an operator of \"integrable\" type in the sense of\nIts-Izergin-Korepin-Slavnov. The construction is not unique and the\nnon-uniqueness highlights the fact that the tau function is really the section\nof a line bundle.\n",
"title": "The Malgrange Form and Fredholm Determinants"
}
| null | null | null | null | true | null |
15904
| null |
Default
| null | null |
null |
{
"abstract": " This paper presents a study of the metaphorism pattern of relational\nspecification, showing how it can be refined into recursive programs.\nMetaphorisms express input-output relationships which preserve relevant\ninformation while at the same time some intended optimization takes place. Text\nprocessing, sorting, representation changers, etc., are examples of\nmetaphorisms. The kind of metaphorism refinement studied in this paper is a\nstrategy known as change of virtual data structure. By framing metaphorisms in\nthe class of (inductive) regular relations, sufficient conditions are given for\nsuch implementations to be calculated using relation algebra. The strategy is\nillustrated with examples including the derivation of the quicksort and\nmergesort algorithms, showing what they have in common and what makes them\ndifferent from the very start of development.\n",
"title": "Programming from Metaphorisms"
}
| null | null | null | null | true | null |
15905
| null |
Default
| null | null |
null |
{
"abstract": " We show that the orthogonal projection operator onto the range of the adjoint\nof a linear operator $T$ can be represented as $UT,$ where $U$ is an invertible\nlinear operator. Using this representation we obtain a decomposition of a\nNormal random vector $Y$ as the sum of a linear transformation of $Y$ that is\nindependent of $TY$ and an affine transformation of $TY$. We then use this\ndecomposition to prove that the conditional distribution of a Normal random\nvector $Y$ given a linear transformation $\\mathcal{T}Y$ is again a multivariate\nNormal distribution. This result is equivalent to the well-known result that\ngiven a $k$-dimensional component of a $n$-dimensional Normal random vector,\nwhere $k<n$, the conditional distribution of the remaining\n$\\left(n-k\\right)$-dimensional component is a $\\left(n-k\\right)$-dimensional\nmultivariate Normal distribution, and sets the stage for approximating the\nconditional distribution of $Y$ given $g\\left(Y\\right)$, where $g$ is a\ncontinuously differentiable vector field.\n",
"title": "On the Conditional Distribution of a Multivariate Normal given a Transformation - the Linear Case"
}
| null | null | null | null | true | null |
15906
| null |
Default
| null | null |
null |
{
"abstract": " We determine the joint limiting distribution of adjacent spacings around a\ncentral, intermediate, or an extreme order statistic $X_{k:n}$ of a random\nsample of size $n$ from a continuous distribution $F$. For central and\nintermediate cases, normalized spacings in the left and right neighborhoods are\nasymptotically i.i.d. exponential random variables. The associated independent\nPoisson arrival processes are independent of $X_{k:n}$. For an extreme\n$X_{k:n}$, the asymptotic independence property of spacings fails for $F$ in\nthe domain of attraction of Fréchet and Weibull ($\\alpha \\neq 1$)\ndistributions. This work also provides additional insight into the limiting\ndistribution for the number of observations around $X_{k:n}$ for all three\ncases.\n",
"title": "Spacings Around An Order Statistic"
}
| null | null | null | null | true | null |
15907
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we prove some fundamental theorems for holomorphic curves on\nangular domain intersecting a hypersurface, finite set of fixed hyperplanes in\ngeneral position and finite set of fixed hypersurfaces in general position on\ncomplex projective variety with the level of truncation. As applications of the\nsecond main theorems for an angle, we will discuss the uniqueness problem of\nholomorphic curves in an angle instead of the whole complex plane. Detail, we\nestablish a result for uniqueness problem of holomorphic curve by inverse image\nof a hypersurface. In my knowledge, this is the first result for uniqueness\nproblem of holomorphic curve by inverse image of hypersurface on angular\ndomain. On complex plane, we obtain a uniqueness result for holomorphic curves,\nit is improvement of some results before [5, 10] in this trend.\n",
"title": "On Nevanlinna - Cartan theory for holomorphic curves with Tsuji characteristics"
}
| null | null | null | null | true | null |
15908
| null |
Default
| null | null |
null |
{
"abstract": " We use the LDA+U approach to search for possible ordered ground states of\nLaSrCoO$_4$. We find a staggered arrangement of magnetic multipoles to be\nstable over a broad range of Co $3d$ interaction parameters. This ordered state\ncan be described as a spin-denity-wave-type condensate of $d_{xy} \\otimes\nd_{x^2-y^2}$ excitons carrying spin $S=1$. Further, we construct an effective\nstrong-coupling model, calculate the exciton dispersion and investigate closing\nof the exciton gap, which marks the exciton condensation instability. Comparing\nthe layered LaSrCoO$_4$ with its pseudo cubic analog LaCoO$_3$, we find that\nfor the same interaction parameters the excitonic gap is smaller (possibly\nvanishing) in the layered cobaltite.\n",
"title": "Theoretical investigation of excitonic magnetism in LaSrCoO$_{4}$"
}
| null | null | null | null | true | null |
15909
| null |
Default
| null | null |
null |
{
"abstract": " In the last decades a vaste amount of evidence for the existence of dark\nmatter has been accumulated. At the same time, many efforts have been\nundertaken to try to identify what dark matter is. Indirect searches look at\nplaces in the Universe where dark matter is believed to be abundant and seek\nfor possible annihilation or decay signatures. The Cherenkov Telescope Array\n(CTA) represents the next generation of imaging Cherenkov telescopes and, with\none site in the Southern hemisphere and one in the Northern hemisphere, will be\nable to observe all the sky with unprecedented sensitivity and angular\nresolution above a few tens of GeV. The CTA Consortium will undertake an\nambitious program of indirect dark matter searches for which we report here the\nbrightest prospects.\n",
"title": "The Dark Matter Programme of the Cherenkov Telescope Array"
}
| null | null | null | null | true | null |
15910
| null |
Default
| null | null |
null |
{
"abstract": " Nuclear starburst discs (NSDs) are star-forming discs that may be residing in\nthe nuclear regions of active galaxies at intermediate redshifts. One\ndimensional (1D) analytical models developed by Thompson et al. (2005) show\nthat these discs can possess an inflationary atmosphere when dust is sublimated\non parsec scales. This make NSDs a viable source for AGN obscuration. We model\nthe two dimensional (2D) structure of NSDs using an iterative method in order\nto compute the explicit vertical solutions for a given annulus. These solutions\nsatisfy energy and hydrostatic balance, as well as the radiative transfer\nequation. In comparison to the 1D model, the 2D calculation predicts a less\nextensive expansion of the atmosphere by orders of magnitude at the\nparsec/sub-parsec scale, but the new scale-height $h$ may still exceed the\nradial distance $R$ for various physical conditions. A total of 192 NSD models\nare computed across the input parameter space in order to predict distributions\nof a line of sight column density $N_H$. Assuming a random distribution of\ninput parameters, the statistics yield 56% of Type 1, 23% of Compton-thin Type\n2s (CN), and 21% of Compton-thick (CK) AGNs. Depending on a viewing angle\n($\\theta$) of a particular NSD (fixed physical conditions), any central AGN can\nappear to be Type 1, CN, or CK which is consistent with the basic unification\ntheory of AGNs. Our results show that $\\log[N_H(\\text{cm}^{-2})]\\in$ [23,25.5]\ncan be oriented at any $\\theta$ from 0$^\\circ$ to $\\approx$80$^\\circ$ due to\nthe degeneracy in the input parameters.\n",
"title": "Modeling the Vertical Structure of Nuclear Starburst Discs: A Possible Source of AGN Obscuration at $z\\sim 1$"
}
| null | null | null | null | true | null |
15911
| null |
Default
| null | null |
null |
{
"abstract": " Despite their vast morphological diversity, many invertebrates have similar\nlarval forms characterized by ciliary bands, innervated arrays of beating cilia\nthat facilitate swimming and feeding. Hydrodynamics suggests that these bands\nshould tightly constrain the behavioral strategies available to the larvae;\nhowever, their apparent ubiquity suggests that these bands also confer\nsubstantial adaptive advantages. Here, we use hydrodynamic techniques to\ninvestigate \"blinking,\" an unusual behavioral phenomenon observed in many\ninvertebrate larvae in which ciliary bands across the body rapidly change\nbeating direction and produce transient rearrangement of the local flow field.\nUsing a general theoretical model combined with quantitative experiments on\nstarfish larvae, we find that the natural rhythm of larval blinking is\nhydrodynamically optimal for inducing strong mixing of the local fluid\nenvironment due to transient streamline crossing, thereby maximizing the\nlarvae's overall feeding rate. Our results are consistent with previous\nhypotheses that filter feeding organisms may use chaotic mixing dynamics to\novercome circulation constraints in viscous environments, and it suggests\nphysical underpinnings for complex neurally-driven behaviors in early-divergent\nanimals.\n",
"title": "Rapid behavioral transitions produce chaotic mixing by a planktonic microswimmer"
}
| null | null | null | null | true | null |
15912
| null |
Default
| null | null |
null |
{
"abstract": " It is well-known that random-coefficient AR(1) process can have long memory\ndepending on the index $\\beta$ of the tail distribution function of the random\ncoefficient, if it is a regularly varying function at unity. We discuss\nestimation of $\\beta$ from panel data comprising N random-coefficient AR(1)\nseries, each of length T. The estimator of $\\beta$ is constructed as a version\nof the tail index estimator of Goldie and Smith (1987) applied to sample lag 1\nautocorrelations of individual time series. Its asymptotic normality is derived\nunder certain conditions on N, T and some parameters of our statistical model.\nBased on this result, we construct a statistical procedure to test if the panel\nrandom-coefficient AR(1) data exhibit long memory. A simulation study\nillustrates finite-sample performance of the introduced estimator and testing\nprocedure.\n",
"title": "Testing for long memory in panel random-coefficient AR(1) data"
}
| null | null | null | null | true | null |
15913
| null |
Default
| null | null |
null |
{
"abstract": " Hilsum-Skandalis maps, from differential geometry, are studied in the context\nof a cartesian category. It is shown that Hilsum-Skandalis maps can be\nrepresented as stably Frobenius adjunctions. This leads to a new and more\ngeneral proof that Hilsum-Skandalis maps represent a universal way of inverting\nessential equivalences between internal groupoids. To prove the representation\ntheorem, a new characterisation of the con- nected components adjunction of any\ninternal groupoid is given. The charaterisation is that the adjunction is\ncovered by a stable Frobenius adjunction that is a slice and whose right\nadjoint is monadic. Geometric morphisms can be represented as stably Frobenius\nadjunctions. As applications of the study we show how it is easy to recover\nproperties of geometric morphisms, seeing them as aspects of properties of\nstably Frobenius adjunctions.\n",
"title": "Hilsum-Skandalis maps as Frobenius adjunctions with application to geometric morphisms"
}
| null | null |
[
"Mathematics"
] | null | true | null |
15914
| null |
Validated
| null | null |
null |
{
"abstract": " We present an explicit version of Berger, Coburn and Lebow's classification\nresult for pure pairs of commuting isometries in the sense of an explicit\nrecipe for constructing pairs of commuting isometric multipliers with precise\ncoefficients. We describe a complete set of (joint) unitary invariants and\ncompare the Berger, Coburn and Lebow's representations with other natural\nanalytic representations of pure pairs of commuting isometries. Finally, we\nstudy the defect operators of pairs of commuting isometries.\n",
"title": "Pairs of commuting isometries - I"
}
| null | null | null | null | true | null |
15915
| null |
Default
| null | null |
null |
{
"abstract": " The paper provides results for the application of boundary feedback control\nwith Zero-Order-Hold (ZOH) to 1-D linear parabolic systems on bounded domains.\nIt is shown that the continuous-time boundary feedback applied in a\nsample-and-hold fashion guarantees closed-loop exponential stability, provided\nthat the sampling period is sufficiently small. Two different continuous-time\nfeedback designs are considered: the reduced model design and the backstepping\ndesign. The obtained results provide stability estimates for weighted 2-norms\nof the state and robustness with respect to perturbations of the sampling\nschedule is guaranteed.\n",
"title": "Sampled-Data Boundary Feedback Control of 1-D Parabolic PDEs"
}
| null | null | null | null | true | null |
15916
| null |
Default
| null | null |
null |
{
"abstract": " Polyethylene Naphtalate (PEN) is a mechanically very favorable polymer.\nEarlier it was found that thin foils made from PEN can have very high\nradio-purity compared to other commercially available foils. In fact, PEN is\nalready in use for low background signal transmission applications (cables).\nRecently it has been realized that PEN also has favorable scintillating\nproperties. In combination, this makes PEN a very promising candidate as a\nself-vetoing structural material in low background experiments. Components\ninstrumented with light detectors could be built from PEN. This includes\ndetector holders, detector containments, signal transmission links, etc. The\ncurrent R\\&D towards qualification of PEN as a self-vetoing low background\nstructural material is be presented.\n",
"title": "PEN as self-vetoing structural Material"
}
| null | null | null | null | true | null |
15917
| null |
Default
| null | null |
null |
{
"abstract": " This is an expanded version of the third author's lecture in String-Math 2015\nat Sanya. It summarizes some of our works in quantum cohomology.\nAfter reviewing the quantum Lefschetz and quantum Leray--Hirsch, we discuss\ntheir applications to the functoriality properties under special smooth flops,\nflips and blow-ups. Finally, for conifold transitions of Calabi--Yau 3-folds,\nformulations for small resolutions (blow-ups along Weil divisors) are sketched.\n",
"title": "Quantum Cohomology under Birational Maps and Transitions"
}
| null | null | null | null | true | null |
15918
| null |
Default
| null | null |
null |
{
"abstract": " We consider the prehomogeneous vector space of pairs of ternary quadratic\nforms. For the lattice of pairs of integral ternary quadratic forms and its\ndual lattice, there are six zeta functions associated with the the\nprehomogeneous vector space. We present a conjecture which states that there\nare simple relations among the six zeta functions. We prove that the\ncoefficients coincide on fundamental discriminants.\n",
"title": "A conjecture on the zeta functions of pairs of ternary quadratic forms"
}
| null | null |
[
"Mathematics"
] | null | true | null |
15919
| null |
Validated
| null | null |
null |
{
"abstract": " Tasks like code generation and semantic parsing require mapping unstructured\n(or partially structured) inputs to well-formed, executable outputs. We\nintroduce abstract syntax networks, a modeling framework for these problems.\nThe outputs are represented as abstract syntax trees (ASTs) and constructed by\na decoder with a dynamically-determined modular structure paralleling the\nstructure of the output tree. On the benchmark Hearthstone dataset for code\ngeneration, our model obtains 79.2 BLEU and 22.7% exact match accuracy,\ncompared to previous state-of-the-art values of 67.1 and 6.1%. Furthermore, we\nperform competitively on the Atis, Jobs, and Geo semantic parsing datasets with\nno task-specific engineering.\n",
"title": "Abstract Syntax Networks for Code Generation and Semantic Parsing"
}
| null | null | null | null | true | null |
15920
| null |
Default
| null | null |
null |
{
"abstract": " The derivation of approximate wave functions for an electron submitted to\nboth a coulomb and a time-dependent laser electric fields, the so-called\nCoulomb-Volkov (CV) state, is addressed. Despite its derivation for continuum\nstates does not exhibit any particular problem within the framework of the\nstandard theory of quantum mechanics (QM), difficulties arise when considering\nan initially bound atomic state. Indeed the natural way of translating the\nunperturbed momentum by the laser vector potential is no longer possible since\na bound state does not exhibit a plane wave form including explicitely a\nmomentum. The use of a fractal space permits to naturally define a momentum for\na bound wave function. Within this framework, it is shown how the derivation of\nlaser-dressed bound states can be performed. Based on a generalized eikonal\napproach, a new expression for the laser-dressed states is also derived, fully\nsymmetric relative to the continuum or bound nature of the initial unperturbed\nwave function. It includes an additional crossed term in the Volkov phase which\nwas not obtained within the standard theory of quantum mechanics. The\nderivations within this fractal framework have highlighted other possible ways\nto derive approximate laser-dressed states in QM. After comparing the various\nobtained wave functions, an application to the prediction of the ionization\nprobability of hydrogen targets by attosecond XUV pulses within the sudden\napproximation is provided. This approach allows to make predictions in various\nregimes depending on the laser intensity, going from the non-resonant\nmultiphoton absorption to tunneling and barrier-suppression ionization.\n",
"title": "Theoretical derivation of laser-dressed atomic states by using a fractal space"
}
| null | null | null | null | true | null |
15921
| null |
Default
| null | null |
null |
{
"abstract": " 'Oumuamua, the first bona-fide interstellar planetesimal, was discovered\npassing through our Solar System on a hyperbolic orbit. This object was likely\ndynamically ejected from an extrasolar planetary system after a series of close\nencounters with gas giant planets. To account for 'Oumuamua's detection, simple\narguments suggest that ~1 Earth mass of planetesimals are ejected per Solar\nmass of Galactic stars. However, that value assumes mono-sized planetesimals.\nIf the planetesimal mass distribution is instead top-heavy the inferred mass in\ninterstellar planetesimals increases to an implausibly high value. The tension\nbetween theoretical expectations for the planetesimal mass function and the\nobservation of 'Oumuamua can be relieved if a small fraction (~0.1-1%) of\nplanetesimals are tidally disrupted on the pathway to ejection into\n'Oumuamua-sized fragments. Using a large suite of simulations of giant planet\ndynamics including planetesimals, we confirm that 0.1-1% of planetesimals pass\nwithin the tidal disruption radius of a gas giant on their pathway to ejection.\n'Oumuamua may thus represent a surviving fragment of a disrupted planetesimal.\nFinally, we argue that an asteroidal composition is dynamically disfavoured for\n'Oumuamua, as asteroidal planetesimals are both less abundant and ejected at a\nlower efficiency than cometary planetesimals.\n",
"title": "Implications of the interstellar object 1I/'Oumuamua for planetary dynamics and planetesimal formation"
}
| null | null |
[
"Physics"
] | null | true | null |
15922
| null |
Validated
| null | null |
null |
{
"abstract": " The implementation of discontinuous Galerkin finite element methods (DGFEMs)\nrepresents a very challenging computational task, particularly for systems of\ncoupled nonlinear PDEs, including multiphysics problems, whose parameters may\nconsist of power series or functionals of the solution variables. Thereby, the\nexploitation of symbolic algebra to express a given DGFEM approximation of a\nPDE problem within a high level language, whose syntax closely resembles the\nmathematical definition, is an invaluable tool. Indeed, this then facilitates\nthe automatic assembly of the resulting system of (nonlinear) equations, as\nwell as the computation of Fréchet derivative(s) of the DGFEM scheme, needed,\nfor example, within a Newton-type solver. However, even exploiting symbolic\nalgebra, the discretisation of coupled systems of PDEs can still be extremely\nverbose and hard to debug. Thereby, in this article we develop a further layer\nof abstraction by designing a class structure for the automatic computation of\nDGFEM formulations. This work has been implemented within the FEniCS package,\nbased on exploiting the Unified Form Language. Numerical examples are presented\nwhich highlight the simplicity of implementation of DGFEMs for the numerical\napproximation of a range of PDE problems.\n",
"title": "Automatic symbolic computation for discontinuous Galerkin finite element methods"
}
| null | null | null | null | true | null |
15923
| null |
Default
| null | null |
null |
{
"abstract": " The quantity and distribution of land which is eligible for renewable energy\nsources is fundamental to the role these technologies will play in future\nenergy systems. As it stands, however, the current state of land eligibility\ninvestigation is found to be insufficient to meet the demands of the future\nenergy modelling community. Three key areas are identified as the predominate\ncauses of this; inconsistent criteria definitions, inconsistent or unclear\nmethodologies, and inconsistent dataset usage. To combat these issues, a land\neligibility framework is developed and described in detail. The validity of\nthis framework is then shown via the recreation of land eligibility results\nfound in the literature, showing strong agreement in the majority of cases.\nFollowing this, the framework is used to perform an evaluation of land\neligibility criteria within the European context whereby the relative\nimportance of commonly considered criteria are compared.\n",
"title": "Methodological Framework for Determining the Land Eligibility of Renewable Energy Sources"
}
| null | null | null | null | true | null |
15924
| null |
Default
| null | null |
null |
{
"abstract": " We conduct an extensive empirical study on short-term electricity price\nforecasting (EPF) to address the long-standing question if the optimal model\nstructure for EPF is univariate or multivariate. We provide evidence that\ndespite a minor edge in predictive performance overall, the multivariate\nmodeling framework does not uniformly outperform the univariate one across all\n12 considered datasets, seasons of the year or hours of the day, and at times\nis outperformed by the latter. This is an indication that combining advanced\nstructures or the corresponding forecasts from both modeling approaches can\nbring a further improvement in forecasting accuracy. We show that this indeed\ncan be the case, even for a simple averaging scheme involving only two models.\nFinally, we also analyze variable selection for the best performing\nhigh-dimensional lasso-type models, thus provide guidelines to structuring\nbetter performing forecasting model designs.\n",
"title": "Day-ahead electricity price forecasting with high-dimensional structures: Univariate vs. multivariate modeling frameworks"
}
| null | null | null | null | true | null |
15925
| null |
Default
| null | null |
null |
{
"abstract": " Let $H$ be a subgroup of the fundamental group $\\pi_{1}(X,x_{0})$. By\nextending the concept of strong SLT space to a relative version with respect to\n$H$, strong $H$-SLT space, first, we investigate the existence of a covering\nmap for strong $H$-SLT spaces. Moreover, we show that a semicovering map is a\ncovering map in the presence of strong $H$-SLT property. Second, we present\nconditions under which the whisker topology agrees with the lasso topology on\n$\\widetilde{X}_{H}$. Also, we study the relationship between open subsets of\n$\\pi_{1}^{wh}(X,x_{0})$ and $\\pi_{1}^{l}(X,x_{0})$. Finally, we give some\nexamples to justify the definition and study of strong $H$-SLT spaces.\n",
"title": "On Strong Small Loop Transfer Spaces Relative to Subgroups of Fundamental Groups"
}
| null | null | null | null | true | null |
15926
| null |
Default
| null | null |
null |
{
"abstract": " We analyze a dataset providing the complete information on the effective\nplays of thousands of music listeners during several months. Our analysis\nconfirms a number of properties previously highlighted by research based on\ninterviews and questionnaires, but also uncover new statistical patterns, both\nat the individual and collective levels. In particular, we show that\nindividuals follow common listening rhythms characterized by the same\nfluctuations, alternating heavy and light listening periods, and can be\nclassified in four groups of similar sizes according to their temporal habits\n--- 'early birds', 'working hours listeners', 'evening listeners' and 'night\nowls'. We provide a detailed radioscopy of the listeners' interplay between\nrepeated listening and discovery of new content. We show that different genres\nencourage different listening habits, from Classical or Jazz music with a more\nbalanced listening among different songs, to Hip Hop and Dance with a more\nheterogeneous distribution of plays. Finally, we provide measures of how\ndistant people are from each other in terms of common songs. In particular, we\nshow that the number of songs $S$ a DJ should play to a random audience of size\n$N$ such that everyone hears at least one song he/she currently listens to, is\nof the form $S\\sim N^\\alpha$ where the exponent depends on the music genre and\nis in the range $[0.5,0.8]$. More generally, our results show that the recent\naccess to virtually infinite catalogs of songs does not promote exploration for\nnovelty, but that most users favor repetition of the same songs.\n",
"title": "Headphones on the wire"
}
| null | null | null | null | true | null |
15927
| null |
Default
| null | null |
null |
{
"abstract": " We present a systematical study via scanning tunneling microscopy (STM) and\nlow-energy electron diffraction (LEED) on the effect of the exposure of Lithium\n(Li) on graphene on silicon carbide (SiC). We have investigated Li deposition\nboth on epitaxial monolayer graphene and on buffer layer surfaces on the\nSi-face of SiC. At room temperature, Li immediately intercalates at the\ninterface between the SiC substrate and the buffer layer and transforms the\nbuffer layer into a quasi-free-standing graphene. This conclusion is\nsubstantiated by LEED and STM evidence. We show that intercalation occurs\nthrough the SiC step sites or graphene defects. We obtain a good quantitative\nagreement between the number of Li atoms deposited and the number of available\nSi bonds at the surface of the SiC crystal. Through STM analysis, we are able\nto determine the interlayer distance induced by Li-intercalation at the\ninterface between the SiC substrate and the buffer layer.\n",
"title": "Li-intercalated Graphene on SiC(0001): an STM study"
}
| null | null | null | null | true | null |
15928
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we propose novel generative models for creating adversarial\nexamples, slightly perturbed images resembling natural images but maliciously\ncrafted to fool pre-trained models. We present trainable deep neural networks\nfor transforming images to adversarial perturbations. Our proposed models can\nproduce image-agnostic and image-dependent perturbations for both targeted and\nnon-targeted attacks. We also demonstrate that similar architectures can\nachieve impressive results in fooling classification and semantic segmentation\nmodels, obviating the need for hand-crafting attack methods for each task.\nUsing extensive experiments on challenging high-resolution datasets such as\nImageNet and Cityscapes, we show that our perturbations achieve high fooling\nrates with small perturbation norms. Moreover, our attacks are considerably\nfaster than current iterative methods at inference time.\n",
"title": "Generative Adversarial Perturbations"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
15929
| null |
Validated
| null | null |
null |
{
"abstract": " We investigate a mixed 0-1 conic quadratic optimization problem with\nindicator variables arising in mean-risk optimization. The indicator variables\nare often used to model non-convexities such as fixed charges or cardinality\nconstraints. Observing that the problem reduces to a submodular function\nminimization for its binary restriction, we derive three classes of strong\nconvex valid inequalities by lifting the polymatroid inequalities on the binary\nvariables. Computational experiments demonstrate the effectiveness of the\ninequalities in strengthening the convex relaxations and, thereby, improving\nthe solution times for mean-risk problems with fixed charges and cardinality\nconstraints significantly.\n",
"title": "Lifted Polymatroid Inequalities for Mean-Risk Optimization with Indicator Variables"
}
| null | null | null | null | true | null |
15930
| null |
Default
| null | null |
null |
{
"abstract": " The atmospheres of exoplanets reveal all their properties beyond mass,\nradius, and orbit. Based on bulk densities, we know that exoplanets larger than\n1.5 Earth radii must have gaseous envelopes, hence atmospheres. We discuss\ncontemporary techniques for characterization of exoplanetary atmospheres. The\nmeasurements are difficult, because - even in current favorable cases - the\nsignals can be as small as 0.001-percent of the host star's flux. Consequently,\nsome early results have been illusory, and not confirmed by subsequent\ninvestigations. Prominent illusions to date include polarized scattered light,\ntemperature inversions, and the existence of carbon planets. The field moves\nfrom the first tentative and often incorrect conclusions, converging to the\nreality of exoplanetary atmospheres. That reality is revealed using transits\nfor close-in exoplanets, and direct imaging for young or massive exoplanets in\ndistant orbits. Several atomic and molecular constituents have now been\nrobustly detected in exoplanets as small as Neptune. In our current\nobservations, the effects of clouds and haze appear ubiquitous. Topics at the\ncurrent frontier include the measurement of heavy element abundances in giant\nplanets, detection of carbon-based molecules, measurement of atmospheric\ntemperature profiles, definition of heat circulation efficiencies for tidally\nlocked planets, and the push to detect and characterize the atmospheres of\nsuper-Earths. Future observatories for this quest include the James Webb Space\nTelescope, and the new generation of Extremely Large Telescopes on the ground.\nOn a more distant horizon, NASA's concepts for the HabEx and LUVOIR missions\ncould extend the study of exoplanetary atmospheres to true twins of Earth.\n",
"title": "Illusion and Reality in the Atmospheres of Exoplanets"
}
| null | null | null | null | true | null |
15931
| null |
Default
| null | null |
null |
{
"abstract": " In concurrent systems, some form of synchronisation is typically needed to\nachieve data-race freedom, which is important for correctness and safety. In\nactor-based systems, messages are exchanged concurrently but executed\nsequentially by the receiving actor. By relying on isolation and non-sharing,\nan actor can access its own state without fear of data-races, and the internal\nbehavior of an actor can be reasoned about sequentially.\nHowever, actor isolation is sometimes too strong to express useful patterns.\nFor example, letting the iterator of a data-collection alias the internal\nstructure of the collection allows a more efficient implementation than if each\naccess requires going through the interface of the collection. With full\nisolation, in order to maintain sequential reasoning the iterator must be made\npart of the collection, which bloats the interface of the collection and means\nthat a client must have access to the whole data-collection in order to use the\niterator.\nIn this paper, we propose a programming language construct that enables a\nrelaxation of isolation but without sacrificing sequential reasoning. We\nformalise the mechanism in a simple lambda calculus with actors and passive\nobjects, and show how an actor may leak parts of its internal state while\nensuring that any interaction with this data is still synchronised.\n",
"title": "Actors without Borders: Amnesty for Imprisoned State"
}
| null | null | null | null | true | null |
15932
| null |
Default
| null | null |
null |
{
"abstract": " Gaussian graphical models are used for determining conditional relationships\nbetween variables. This is accomplished by identifying off-diagonal elements in\nthe inverse-covariance matrix that are non-zero. When the ratio of variables\n(p) to observations (n) approaches one, the maximum likelihood estimator of the\ncovariance matrix becomes unstable and requires shrinkage estimation. Whereas\nseveral classical (frequentist) methods have been introduced to address this\nissue, fully Bayesian methods remain relatively uncommon in practice and\nmethodological literatures. Here we introduce a Bayesian method for estimating\nsparse matrices, in which conditional relationships are determined with\nprojection predictive selection. With this method, that uses Kullback-Leibler\ndivergence and cross-validation for neighborhood selection, we reconstruct the\ninverse-covariance matrix in both low and high-dimensional settings. Through\nsimulation and applied examples, we characterized performance compared to\nseveral Bayesian methods and the graphical lasso, in addition to TIGER that\nsimilarly estimates the inverse-covariance matrix with regression. Our results\ndemonstrate that projection predictive selection not only has superior\nperformance compared to selecting the most probable model and Bayesian model\naveraging, particularly for high-dimensional data, but also compared to the the\nBayesian and classical glasso methods. Further, we show that estimating the\ninverse-covariance matrix with multiple regression is often more accurate, with\nrespect to various loss functions, and efficient than direct estimation. In\nlow-dimensional settings, we demonstrate that projection predictive selection\nalso provides competitive performance. We have implemented the projection\npredictive method for covariance selection in the R package GGMprojpred\n",
"title": "Bayesian Estimation of Gaussian Graphical Models with Predictive Covariance Selection"
}
| null | null | null | null | true | null |
15933
| null |
Default
| null | null |
null |
{
"abstract": " While deep learning models have achieved state-of-the-art accuracies for many\nprediction tasks, understanding these models remains a challenge. Despite the\nrecent interest in developing visual tools to help users interpret deep\nlearning models, the complexity and wide variety of models deployed in\nindustry, and the large-scale datasets that they used, pose unique design\nchallenges that are inadequately addressed by existing work. Through\nparticipatory design sessions with over 15 researchers and engineers at\nFacebook, we have developed, deployed, and iteratively improved ActiVis, an\ninteractive visualization system for interpreting large-scale deep learning\nmodels and results. By tightly integrating multiple coordinated views, such as\na computation graph overview of the model architecture, and a neuron activation\nview for pattern discovery and comparison, users can explore complex deep\nneural network models at both the instance- and subset-level. ActiVis has been\ndeployed on Facebook's machine learning platform. We present case studies with\nFacebook researchers and engineers, and usage scenarios of how ActiVis may work\nwith different models.\n",
"title": "ActiVis: Visual Exploration of Industry-Scale Deep Neural Network Models"
}
| null | null | null | null | true | null |
15934
| null |
Default
| null | null |
null |
{
"abstract": " For a unit vector field on a closed immersed Euclidean hypersurface\n$M^{2n+1}$, $n\\geq 1$, we exhibit a nontrivial lower bound for its energy which\ndepends on the degree of the Gauss map of the immersion. When the hypersurface\nis the unit sphere $\\mathbb{S}^{2n+1}$, immersed with degree one, this lower\nbound corresponds to a well established value from the literature. We introduce\na list of functionals $\\mathcal{B}_k$ on a compact Riemannian manifold $M^{m}$,\n$1\\leq k\\leq m$, and show that, when the underlying manifold is a closed\nhypersurface, these functionals possess similar properties regarding the degree\nof the immersion. In addition, we prove that Hopf flows minimize\n$\\mathcal{B}_n$ on $\\mathbb{S}^{2n+1}$.\n",
"title": "A topological lower bound for the energy of a unit vector field on a closed Euclidean hypersurface"
}
| null | null | null | null | true | null |
15935
| null |
Default
| null | null |
null |
{
"abstract": " The Lyapunov rank of a proper cone $K$ in a finite dimensional real Hilbert\nspace is defined as the dimension of the space of all Lyapunov-like\ntransformations on $K$, or equivalently, the dimension of the Lie algebra of\nthe automorphism group of $K$. This (rank) measures the number of linearly\nindependent bilinear relations needed to express a complementarity system on\n$K$ (that arises, for example, from a linear program or a complementarity\nproblem on the cone). Motivated by the problem of describing spectral/proper\ncones where the complementarity system can be expressed as a square system\n(that is, where the Lyapunov rank is greater than equal to the dimension of the\nambient space), we consider proper polyhedral cones in $\\mathbb{R}^n$ that are\npermutation invariant. For such cones we show that the Lyapunov rank is either\n1 (in which case, the cone is irreducible) or n (in which case, the cone is\nisomorphic to the nonnegative orthart in $\\mathbb{R}^n$). In the latter case,\nwe show that the corresponding spectral cone is isomorphic to a symmetric cone.\n",
"title": "Permutation invariant proper polyhedral cones and their Lyapunov rank"
}
| null | null |
[
"Mathematics"
] | null | true | null |
15936
| null |
Validated
| null | null |
null |
{
"abstract": " Detecting feature interactions is imperative for accurately predicting\nperformance of highly-configurable systems. State-of-the-art performance\nprediction techniques rely on supervised machine learning for detecting feature\ninteractions, which, in turn, relies on time consuming performance measurements\nto obtain training data. By providing information about potentially interacting\nfeatures, we can reduce the number of required performance measurements and\nmake the overall performance prediction process more time efficient. We expect\nthat the information about potentially interacting features can be obtained by\nstatically analyzing the source code of a highly-configurable system, which is\ncomputationally cheaper than performing multiple performance measurements. To\nthis end, we conducted a qualitative case study in which we explored the\nrelation between control-flow feature interactions (detected through static\nprogram analysis) and performance feature interactions (detected by performance\nprediction techniques using performance measurements). We found that a relation\nexists, which can potentially be exploited to predict performance interactions.\n",
"title": "On the Relation of External and Internal Feature Interactions: A Case Study"
}
| null | null | null | null | true | null |
15937
| null |
Default
| null | null |
null |
{
"abstract": " Molecular adsorption on surfaces plays an important part in catalysis,\ncorrosion, desalination, and various other processes that are relevant to\nindustry and in nature. As a complement to experiments, accurate adsorption\nenergies can be obtained using various sophisticated electronic structure\nmethods that can now be applied to periodic systems. The adsorption energy of\nwater on boron nitride substrates, going from zero to 2-dimensional\nperiodicity, is particularly interesting as it calls for an accurate treatment\nof polarizable electrostatics and dispersion interactions, as well as posing a\npractical challenge to experiments and electronic structure methods. Here, we\npresent reference adsorption energies, static polarizabilities, and dynamic\npolarizabilities, for water on BN substrates of varying size and dimension.\nAdsorption energies are computed with coupled cluster theory, fixed-node\nquantum Monte Carlo (FNQMC), the random phase approximation (RPA), and second\norder M{\\o}ller-Plesset (MP2) theory. These explicitly correlated methods are\nfound to agree in molecular as well as periodic systems. The best estimate of\nthe water/h-BN adsorption energy is $-107\\pm7$ meV from FNQMC. In addition, the\nwater adsorption energy on the BN substrates could be expected to grow\nmonotonically with the size of the substrate due to increased dispersion\ninteractions but interestingly, this is not the case here. This peculiar\nfinding is explained using the static polarizabilities and molecular dispersion\ncoefficients of the systems, as computed from time-dependent density functional\ntheory (DFT). Dynamic as well as static polarizabilities are found to be highly\nanisotropic in these systems. In addition, the many-body dispersion method in\nDFT emerges as a particularly useful estimation of finite size effects for\nother expensive, many-body wavefunction based methods.\n",
"title": "Properties of the water to boron nitride interaction: from zero to two dimensions with benchmark accuracy"
}
| null | null | null | null | true | null |
15938
| null |
Default
| null | null |
null |
{
"abstract": " We theoretically investigate the mechanism to generate large intrinsic spin\nHall effect in iridates or more broadly in 5d transition metal oxides with\nstrong spin-orbit coupling. We demonstrate such a possibility by taking the\nexample of orthorhombic perovskite iridate with nonsymmorphic lattice symmetry,\nSrIrO$_3$, which is a three-dimensional semimetal with nodal line spectrum. It\nis shown that large intrinsic spin Hall effect arises in this system via the\nspin-Berry curvature originating from the nearly degenerate electronic spectra\nsurrounding the nodal line. This effect exists even when the nodal line is\ngently gapped out, due to the persistent nearly degenerate electronic\nstructure, suggesting a distinct robustness. The magnitude of the spin Hall\nconductivity is shown to be comparable to the best known example such as doped\ntopological insulators and the biggest in any transition metal oxides. To gain\nfurther insight, we compute the intrinsic spin Hall conductivity in both of the\nbulk and thin film systems. We find that the geometric confinement in thin\nfilms leads to significant modifications of the electronic states, leading to\neven bigger spin Hall conductivity in certain cases. We compare our findings\nwith the recent experimental report on the discovery of large spin Hall effect\nin SrIrO$_3$ thin films.\n",
"title": "Theory of Large Intrinsic Spin Hall Effect in Iridate Semimetals"
}
| null | null | null | null | true | null |
15939
| null |
Default
| null | null |
null |
{
"abstract": " The analysis of cancer genomic data has long suffered \"the curse of\ndimensionality\". Sample sizes for most cancer genomic studies are a few\nhundreds at most while there are tens of thousands of genomic features studied.\nVarious methods have been proposed to leverage prior biological knowledge, such\nas pathways, to more effectively analyze cancer genomic data. Most of the\nmethods focus on testing marginal significance of the associations between\npathways and clinical phenotypes. They can identify relevant pathways, but do\nnot involve predictive modeling. In this article, we propose a Pathway-based\nKernel Boosting (PKB) method for integrating gene pathway information for\nsample classification, where we use kernel functions calculated from each\npathway as base learners and learn the weights through iterative optimization\nof the classification loss function. We apply PKB and several competing methods\nto three cancer studies with pathological and clinical information, including\ntumor grade, stage, tumor sites, and metastasis status. Our results show that\nPKB outperforms other methods, and identifies pathways relevant to the outcome\nvariables.\n",
"title": "A pathway-based kernel boosting method for sample classification using genomic data"
}
| null | null | null | null | true | null |
15940
| null |
Default
| null | null |
null |
{
"abstract": " The graph Fourier transform (GFT) is in general dense and requires O(n^2)\ntime to compute and O(n^2) memory space to store. In this paper, we pursue our\nprevious work on the approximate fast graph Fourier transform (FGFT). The FGFT\nis computed via a truncated Jacobi algorithm, and is defined as the product of\nJ Givens rotations (very sparse orthogonal matrices). The truncation parameter,\nJ, represents a trade-off between precision of the transform and time of\ncomputation (and storage space). We explore further this trade-off and study,\non different types of graphs, how is the approximation error distributed along\nthe spectrum.\n",
"title": "Analyzing the Approximation Error of the Fast Graph Fourier Transform"
}
| null | null | null | null | true | null |
15941
| null |
Default
| null | null |
null |
{
"abstract": " Neural networks are capable of learning rich, nonlinear feature\nrepresentations shown to be beneficial in many predictive tasks. In this work,\nwe use such models to explore different geographical feature representations in\nthe context of predicting colorectal cancer survival curves for patients in the\nstate of Iowa, spanning the years 1989 to 2013. Specifically, we compare model\nperformance using \"area between the curves\" (ABC) to assess (a) whether\nsurvival curves can be reasonably predicted for colorectal cancer patients in\nthe state of Iowa, (b) whether geographical features improve predictive\nperformance, (c) whether a simple binary representation, or a richer, spectral\nanalysis-elicited representation perform better, and (d) whether spectral\nanalysis-based representations can be improved upon by leveraging\ngeographically-descriptive features. In exploring (d), we devise a\nsimilarity-based spectral analysis procedure, which allows for the combination\nof geographically relational and geographically descriptive features. Our\nfindings suggest that survival curves can be reasonably estimated on average,\nwith predictive performance deviating at the five-year survival mark among all\nmodels. We also find that geographical features improve predictive performance,\nand that better performance is obtained using richer, spectral\nanalysis-elicited features. Furthermore, we find that similarity-based spectral\nanalysis-elicited representations improve upon the original spectral analysis\nresults by approximately 40%.\n",
"title": "Deriving Enhanced Geographical Representations via Similarity-based Spectral Analysis: Predicting Colorectal Cancer Survival Curves in Iowa"
}
| null | null |
[
"Statistics"
] | null | true | null |
15942
| null |
Validated
| null | null |
null |
{
"abstract": " In the prize-collecting Steiner forest (PCSF) problem, we are given an\nundirected graph $G=(V,E)$, edge costs $\\{c_e\\geq 0\\}_{e\\in E}$, terminal pairs\n$\\{(s_i,t_i)\\}_{i=1}^k$, and penalties $\\{\\pi_i\\}_{i=1}^k$ for each terminal\npair; the goal is to find a forest $F$ to minimize $c(F)+\\sum_{i:\n(s_i,t_i)\\text{ not connected in }F}\\pi_i$. The Steiner forest problem can be\nviewed as the special case where $\\pi_i=\\infty$ for all $i$. It was widely\nbelieved that the integrality gap of the natural (and well-studied)\nlinear-programming (LP) relaxation for PCSF is at most 2. We dispel this belief\nby showing that the integrality gap of this LP is at least $9/4$. This holds\neven for planar graphs. We also show that using this LP, one cannot devise a\nLagrangian-multiplier-preserving (LMP) algorithm with approximation guarantee\nbetter than $4$. Our results thus show a separation between the integrality\ngaps of the LP-relaxations for prize-collecting and non-prize-collecting (i.e.,\nstandard) Steiner forest, as well as the approximation ratios achievable\nrelative to the optimal LP solution by LMP- and non-LMP- approximation\nalgorithms for PCSF. For the special case of prize-collecting Steiner tree\n(PCST), we prove that the natural LP relaxation admits basic feasible solutions\nwith all coordinates of value at most $1/3$ and all edge variables positive.\nThus, we rule out the possibility of approximating PCST with guarantee better\nthan $3$ using a direct iterative rounding method.\n",
"title": "On the Integrality Gap of the Prize-Collecting Steiner Forest LP"
}
| null | null | null | null | true | null |
15943
| null |
Default
| null | null |
null |
{
"abstract": " There is a large literature on semiparametric estimation of average treatment\neffects under unconfounded treatment assignment in settings with a fixed number\nof covariates. More recently attention has focused on settings with a large\nnumber of covariates. In this paper we extend lessons from the earlier\nliterature to this new setting. We propose that in addition to reporting point\nestimates and standard errors, researchers report results from a number of\nsupplementary analyses to assist in assessing the credibility of their\nestimates.\n",
"title": "Estimating Average Treatment Effects: Supplementary Analyses and Remaining Challenges"
}
| null | null | null | null | true | null |
15944
| null |
Default
| null | null |
null |
{
"abstract": " We study discretizations of polynomial processes using finite state Markov\nprocesses satisfying suitable moment matching conditions. The states of these\nMarkov processes together with their transition probabilities can be\ninterpreted as Markov cubature rules. The polynomial property allows us to\nstudy such rules using algebraic techniques. Markov cubature rules aid the\ntractability of path-dependent tasks such as American option pricing in models\nwhere the underlying factors are polynomial processes.\n",
"title": "Markov cubature rules for polynomial processes"
}
| null | null | null | null | true | null |
15945
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we propose a new method for estimation and constructing\nconfidence intervals for low-dimensional components in a high-dimensional\nmodel. The proposed estimator, called Constrained Lasso (CLasso) estimator, is\nobtained by simultaneously solving two estimating equations---one imposing a\nzero-bias constraint for the low-dimensional parameter and the other forming an\n$\\ell_1$-penalized procedure for the high-dimensional nuisance parameter. By\ncarefully choosing the zero-bias constraint, the resulting estimator of the low\ndimensional parameter is shown to admit an asymptotically normal limit\nattaining the Cramér-Rao lower bound in a semiparametric sense. We propose\na tuning-free iterative algorithm for implementing the CLasso. We show that\nwhen the algorithm is initialized at the Lasso estimator, the de-sparsified\nestimator proposed in van de Geer et al. [\\emph{Ann. Statist.} {\\bf 42} (2014)\n1166--1202] is asymptotically equivalent to the first iterate of the algorithm.\nWe analyse the asymptotic properties of the CLasso estimator and show the\nglobally linear convergence of the algorithm. We also demonstrate encouraging\nempirical performance of the CLasso through numerical studies.\n",
"title": "Statistical inference for high dimensional regression via Constrained Lasso"
}
| null | null | null | null | true | null |
15946
| null |
Default
| null | null |
null |
{
"abstract": " We present an extragalactic survey using observations from the Atacama Large\nMillimeter/submillimeter Array (ALMA) to characterise galaxy populations up to\n$z=0.35$: the Valparaíso ALMA Line Emission Survey (VALES). We use ALMA\nBand-3 CO(1--0) observations to study the molecular gas content in a sample of\n67 dusty normal star-forming galaxies selected from the $Herschel$\nAstrophysical Terahertz Large Area Survey ($H$-ATLAS). We have spectrally\ndetected 49 galaxies at $>5\\sigma$ significance and 12 others are seen at low\nsignificance in stacked spectra. CO luminosities are in the range of\n$(0.03-1.31)\\times10^{10}$ K km s$^{-1}$ pc$^2$, equivalent to $\\log({\\rm\nM_{gas}/M_{\\odot}}) =8.9-10.9$ assuming an $\\alpha_{\\rm CO}$=4.6(K km s$^{-1}$\npc$^{2}$)$^{-1}$, which perfectly complements the parameter space previously\nexplored with local and high-z normal galaxies. We compute the optical to CO\nsize ratio for 21 galaxies resolved by ALMA at $\\sim 3$.\"$5$ resolution (6.5\nkpc), finding that the molecular gas is on average $\\sim$ 0.6 times more\ncompact than the stellar component. We obtain a global Schmidt-Kennicutt\nrelation, given by $\\log [\\Sigma_{\\rm SFR}/({\\rm M_{\\odot}\nyr^{-1}kpc^{-2}})]=(1.26 \\pm 0.02) \\times \\log [\\Sigma_{\\rm M_{H2}}/({\\rm\nM_{\\odot}\\,pc^{-2}})]-(3.6 \\pm 0.2)$. We find a significant fraction of\ngalaxies lying at `intermediate efficiencies' between a long-standing mode of\nstar-formation activity and a starburst, specially at $\\rm L_{IR}=10^{11-12}\nL_{\\odot}$. Combining our observations with data taken from the literature, we\npropose that star formation efficiencies can be parameterised by $\\log [{\\rm\nSFR/M_{H2}}]=0.19 \\times {\\rm (\\log {L_{IR}}-11.45)}-8.26-0.41 \\times\n\\arctan[-4.84 (\\log {\\rm L_{IR}}-11.45) ]$. Within the redshift range we\nexplore ($z<0.35$), we identify a rapid increase of the gas content as a\nfunction of redshift.\n",
"title": "VALES: I. The molecular gas content in star-forming dusty H-ATLAS galaxies up to z=0.35"
}
| null | null | null | null | true | null |
15947
| null |
Default
| null | null |
null |
{
"abstract": " For a degenerate autonomous Kirchhoff equation which is set on $\\mathbb{R}^N$\nand involves the Berestycki-Lions type nonlinearity, we cope with the cases\n$N=2,3$ and $N\\geq5$ by using mountain pass and symmetric mountain pass\napproaches and by using Clark theorem respectively.\n",
"title": "Variational methods for degenerate Kirchhoff equations"
}
| null | null | null | null | true | null |
15948
| null |
Default
| null | null |
null |
{
"abstract": " The Doppler tracking data of the Chang'e 3 lunar mission is used to constrain\nthe stochastic background of gravitational wave in cosmology within the 1 mHz\nto 0.05 Hz frequency band. Our result improves on the upper bound on the energy\ndensity of the stochastic background of gravitational wave in the 0.02 Hz to\n0.05 Hz band obtained by the Apollo missions, with the improvement reaching\nalmost one order of magnitude at around 0.05 Hz. Detailed noise analysis of the\nDoppler tracking data is also presented, with the prospect that these noise\nsources will be mitigated in future Chinese deep space missions. A feasibility\nstudy is also undertaken to understand the scientific capability of the Chang'e\n4 mission, due to be launched in 2018, in relation to the stochastic\ngravitational wave background around 0.01 Hz. The study indicates that the\nupper bound on the energy density may be further improved by another order of\nmagnitude from the Chang'e 3 mission, which will fill the gap in the frequency\nband from 0.02 Hz to 0.1 Hz in the foreseeable future.\n",
"title": "Chang'e 3 lunar mission and upper limit on stochastic background of gravitational wave around the 0.01 Hz band"
}
| null | null | null | null | true | null |
15949
| null |
Default
| null | null |
null |
{
"abstract": " Modern tracking technology has made the collection of large numbers of\ndensely sampled trajectories of moving objects widely available. We consider a\nfundamental problem encountered when analysing such data: Given $n$ polygonal\ncurves $S$ in $\\mathbb{R}^d$, preprocess $S$ into a data structure that answers\nqueries with a query curve $q$ and radius $\\rho$ for the curves of $S$ that\nhave \\Frechet distance at most $\\rho$ to $q$.\nWe initiate a comprehensive analysis of the space/query-time trade-off for\nthis data structuring problem. Our lower bounds imply that any data structure\nin the pointer model model that achieves $Q(n) + O(k)$ query time, where $k$ is\nthe output size, has to use roughly $\\Omega\\left((n/Q(n))^2\\right)$ space in\nthe worst case, even if queries are mere points (for the discrete \\Frechet\ndistance) or line segments (for the continuous \\Frechet distance). More\nimportantly, we show that more complex queries and input curves lead to\nadditional logarithmic factors in the lower bound. Roughly speaking, the number\nof logarithmic factors added is linear in the number of edges added to the\nquery and input curve complexity. This means that the space/query time\ntrade-off worsens by an exponential factor of input and query complexity. This\nbehaviour addresses an open question in the range searching literature: whether\nit is possible to avoid the additional logarithmic factors in the space and\nquery time of a multilevel partition tree. We answer this question negatively.\nOn the positive side, we show we can build data structures for the \\Frechet\ndistance by using semialgebraic range searching. Our solution for the discrete\n\\Frechet distance is in line with the lower bound, as the number of levels in\nthe data structure is $O(t)$, where $t$ denotes the maximal number of vertices\nof a curve. For the continuous \\Frechet distance, the number of levels\nincreases to $O(t^2)$.\n",
"title": "On the complexity of range searching among curves"
}
| null | null | null | null | true | null |
15950
| null |
Default
| null | null |
null |
{
"abstract": " We propose a Topic Compositional Neural Language Model (TCNLM), a novel\nmethod designed to simultaneously capture both the global semantic meaning and\nthe local word ordering structure in a document. The TCNLM learns the global\nsemantic coherence of a document via a neural topic model, and the probability\nof each learned latent topic is further used to build a Mixture-of-Experts\n(MoE) language model, where each expert (corresponding to one topic) is a\nrecurrent neural network (RNN) that accounts for learning the local structure\nof a word sequence. In order to train the MoE model efficiently, a matrix\nfactorization method is applied, by extending each weight matrix of the RNN to\nbe an ensemble of topic-dependent weight matrices. The degree to which each\nmember of the ensemble is used is tied to the document-dependent probability of\nthe corresponding topics. Experimental results on several corpora show that the\nproposed approach outperforms both a pure RNN-based model and other\ntopic-guided language models. Further, our model yields sensible topics, and\nalso has the capacity to generate meaningful sentences conditioned on given\ntopics.\n",
"title": "Topic Compositional Neural Language Model"
}
| null | null | null | null | true | null |
15951
| null |
Default
| null | null |
null |
{
"abstract": " We numerically investigate the electronic transport properties of graphene\nnanoribbons and carbon nanotubes with inter-valley coupling, e.g., in \\sqrt{3}N\n\\times \\sqrt{3}N and 3N \\times 3N superlattices. By taking the \\sqrt{3} \\times\n\\sqrt{3} graphene superlattice as an example, we show that tailoring the bulk\ngraphene superlattice results in rich structural configurations of nanoribbons\nand nanotubes. After studying the electronic characteristics of the\ncorresponding armchair and zigzag nanoribbon geometries, we find that the\nlinear bands of carbon nanotubes can lead to the Klein tunnelling-like\nphenomenon, i.e., electrons propagate along tubes without backscattering even\nin the presence of a barrier. Due to the coupling between K and K' valleys of\npristine graphene by \\sqrt{3} \\times \\sqrt{3} supercells,we propose a\nvalley-field-effect transistor based on the armchair carbon nanotube, where the\nvalley polarization of the current can be tuned by applying a gate voltage or\nvarying the length of the armchair carbon nanotubes.\n",
"title": "Transmission spectra and valley processing of graphene and carbon nanotube superlattices with inter-valley coupling"
}
| null | null | null | null | true | null |
15952
| null |
Default
| null | null |
null |
{
"abstract": " Generative modeling of high dimensional data like images is a notoriously\ndifficult and ill-defined problem. In particular, how to evaluate a learned\ngenerative model is unclear. In this position paper, we argue that adversarial\nlearning, pioneered with generative adversarial networks (GANs), provides an\ninteresting framework to implicitly define more meaningful task losses for\ngenerative modeling tasks, such as for generating \"visually realistic\" images.\nWe refer to those task losses as parametric adversarial divergences and we give\ntwo main reasons why we think parametric divergences are good learning\nobjectives for generative modeling. Additionally, we unify the processes of\nchoosing a good structured loss (in structured prediction) and choosing a\ndiscriminator architecture (in generative modeling) using statistical decision\ntheory; we are then able to formalize and quantify the intuition that \"weaker\"\nlosses are easier to learn from, in a specific setting. Finally, we propose two\nnew challenging tasks to evaluate parametric and nonparametric divergences: a\nqualitative task of generating very high-resolution digits, and a quantitative\ntask of learning data that satisfies high-level algebraic constraints. We use\ntwo common divergences to train a generator and show that the parametric\ndivergence outperforms the nonparametric divergence on both the qualitative and\nthe quantitative task.\n",
"title": "Parametric Adversarial Divergences are Good Task Losses for Generative Modeling"
}
| null | null | null | null | true | null |
15953
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we consider a $d$-dimensional ($d=1,2$) parabolic-elliptic\nKeller-Segel equation with a logistic forcing and a fractional diffusion of\norder $\\alpha \\in (0,2)$. We prove uniform in time boundedness of its solution\nin the supercritical range $\\alpha>d\\left(1-c\\right)$, where $c$ is an explicit\nconstant depending on parameters of our problem. Furthermore, we establish\nsufficient conditions for $\\|u(t)-u_\\infty\\|_{L^\\infty}\\rightarrow0$, where\n$u_\\infty\\equiv 1$ is the only nontrivial homogeneous solution. Finally, we\nprovide a uniqueness result.\n",
"title": "Boundedness and homogeneous asymptotics for a fractional logistic Keller-Segel equations"
}
| null | null | null | null | true | null |
15954
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we prove explicit formulas for all Willmore surfaces of\nrevolution and demonstrate their use in the discussion of the associated\nDirichlet boundary value problems. It is shown by an explicit example that\nsymmetric Dirichlet boundary conditions do in general not entail the symmetry\nof the surface. In addition we prove a symmetry result for a subclass of\nWillmore surfaces satisfying symmetric Dirichlet boundary data.\n",
"title": "Explicit formulas, symmetry and symmetry breaking for Willmore surfaces of revolution"
}
| null | null | null | null | true | null |
15955
| null |
Default
| null | null |
null |
{
"abstract": " We consider the problem of optimal budget allocation for crowdsourcing\nproblems, allocating users to tasks to maximize our final confidence in the\ncrowdsourced answers. Such an optimized worker assignment method allows us to\nboost the efficacy of any popular crowdsourcing estimation algorithm. We\nconsider a mutual information interpretation of the crowdsourcing problem,\nwhich leads to a stochastic subset selection problem with a submodular\nobjective function. We present experimental simulation results which\ndemonstrate the effectiveness of our dynamic task allocation method for\nachieving higher accuracy, possibly requiring fewer labels, as well as\nimproving upon a previous method which is sensitive to the proportion of users\nto questions.\n",
"title": "Dynamic Task Allocation for Crowdsourcing Settings"
}
| null | null | null | null | true | null |
15956
| null |
Default
| null | null |
null |
{
"abstract": " We study unsupervised generative modeling in terms of the optimal transport\n(OT) problem between true (but unknown) data distribution $P_X$ and the latent\nvariable model distribution $P_G$. We show that the OT problem can be\nequivalently written in terms of probabilistic encoders, which are constrained\nto match the posterior and prior distributions over the latent space. When\nrelaxed, this constrained optimization problem leads to a penalized optimal\ntransport (POT) objective, which can be efficiently minimized using stochastic\ngradient descent by sampling from $P_X$ and $P_G$. We show that POT for the\n2-Wasserstein distance coincides with the objective heuristically employed in\nadversarial auto-encoders (AAE) (Makhzani et al., 2016), which provides the\nfirst theoretical justification for AAEs known to the authors. We also compare\nPOT to other popular techniques like variational auto-encoders (VAE) (Kingma\nand Welling, 2014). Our theoretical results include (a) a better understanding\nof the commonly observed blurriness of images generated by VAEs, and (b)\nestablishing duality between Wasserstein GAN (Arjovsky and Bottou, 2017) and\nPOT for the 1-Wasserstein distance.\n",
"title": "From optimal transport to generative modeling: the VEGAN cookbook"
}
| null | null | null | null | true | null |
15957
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we study the Gauss map of a free boundary minimal surface. The\nmain theorem asserts that if components of the Gauss map are eigenfunctions of\nthe Jacobi-Steklov operator, then the surface must be rotationally symmetric.\n",
"title": "The Gauss map of a free boundary minimal surface"
}
| null | null |
[
"Mathematics"
] | null | true | null |
15958
| null |
Validated
| null | null |
null |
{
"abstract": " Computational approaches to finding non-trivial integer solutions of the\nequation in the title are discussed. We summarize previous work and provide\nseveral new solutions.\n",
"title": "Computational Experiments on $a^4+b^4+c^4+d^4=(a+b+c+d)^4$"
}
| null | null |
[
"Mathematics"
] | null | true | null |
15959
| null |
Validated
| null | null |
null |
{
"abstract": " We present a variation of the Autoencoder (AE) that explicitly maximizes the\nmutual information between the input data and the hidden representation. The\nproposed model, the InfoMax Autoencoder (IMAE), by construction is able to\nlearn a robust representation and good prototypes of the data. IMAE is compared\nboth theoretically and then computationally with the state of the art models:\nthe Denoising and Contractive Autoencoders in the one-hidden layer setting and\nthe Variational Autoencoder in the multi-layer case. Computational experiments\nare performed with the MNIST and Fashion-MNIST datasets and demonstrate\nparticularly the strong clusterization performance of IMAE.\n",
"title": "An information theoretic approach to the autoencoder"
}
| null | null | null | null | true | null |
15960
| null |
Default
| null | null |
null |
{
"abstract": " Planning motions for two robot arms to move an object collaboratively is a\ndifficult problem, mainly because of the closed-chain constraint, which arises\nwhenever two robot hands simultaneously grasp a single rigid object. In this\npaper, we propose a manipulation planning algorithm to bring an object from an\ninitial stable placement (position and orientation of the object on the support\nsurface) towards a goal stable placement. The key specificity of our algorithm\nis that it is certified-complete: for a given object and a given environment,\nwe provide a certificate that the algorithm will find a solution to any\nbimanual manipulation query in that environment whenever one exists. Moreover,\nthe certificate is constructive: at run-time, it can be used to quickly find a\nsolution to a given query. The algorithm is tested in software and hardware on\na number of large pieces of furniture.\n",
"title": "A Certified-Complete Bimanual Manipulation Planner"
}
| null | null | null | null | true | null |
15961
| null |
Default
| null | null |
null |
{
"abstract": " The Short-Baseline Neutrino (SBN) Program is a short-baseline neutrino\noscillation experiment in the Booster Neutrino Beam-line (BNB) at Fermilab. It\nconsists of three Liquid Argon Time Projection Chambers (LArTPCs) from the\nShort-Baseline Near Detector (SBND), Micro Booster Neutrino Experiment\n(MicroBooNE), and Imaging Cosmic And Rare Underground Signals (ICARUS)\nexperiments. The SBN Program will definitively search for short-baseline\nneutrino oscillations in the 1 eV mass range, make precision neutrino-argon\ninteraction measurements, and further develop the LArTPC technology. The\nphysics program and current status of the program, and its constituent\nexperiments, are presented.\n",
"title": "The Short Baseline Neutrino Oscillation Program at Fermilab"
}
| null | null | null | null | true | null |
15962
| null |
Default
| null | null |
null |
{
"abstract": " A novel algorithm is proposed for CANDECOMP/PARAFAC tensor decomposition to\nexploit best rank-1 tensor approximation. Different from the existing\nalgorithms, our algorithm updates rank-1 tensors simultaneously in parallel. In\norder to achieve this, we develop new all-at-once algorithms for best rank-1\ntensor approximation based on the Levenberg-Marquardt method and the rotational\nupdate. We show that the LM algorithm has the same complexity of first-order\noptimisation algorithms, while the rotational method leads to solving the best\nrank-1 approximation of tensors of size $2 \\times 2 \\times \\cdots \\times 2$. We\nderive a closed-form expression of the best rank-1 tensor of $2\\times 2 \\times\n2$ tensors and present an ALS algorithm which updates 3 component at a time for\nhigher order tensors. The proposed algorithm is illustrated in decomposition of\ndifficult tensors which are associated with multiplication of two matrices.\n",
"title": "Best Rank-One Tensor Approximation and Parallel Update Algorithm for CPD"
}
| null | null | null | null | true | null |
15963
| null |
Default
| null | null |
null |
{
"abstract": " Testing (conditional) independence of multivariate random variables is a task\ncentral to statistical inference and modelling in general - though\nunfortunately one for which to date there does not exist a practicable\nworkflow. State-of-art workflows suffer from the need for heuristic or\nsubjective manual choices, high computational complexity, or strong parametric\nassumptions.\nWe address these problems by establishing a theoretical link between\nmultivariate/conditional independence testing, and model comparison in the\nmultivariate predictive modelling aka supervised learning task. This link\nallows advances in the extensively studied supervised learning workflow to be\ndirectly transferred to independence testing workflows - including automated\ntuning of machine learning type which addresses the need for a heuristic\nchoice, the ability to quantitatively trade-off computational demand with\naccuracy, and the modern black-box philosophy for checking and interfacing.\nAs a practical implementation of this link between the two workflows, we\npresent a python package 'pcit', which implements our novel multivariate and\nconditional independence tests, interfacing the supervised learning API of the\nscikit-learn package. Theory and package also allow for straightforward\nindependence test based learning of graphical model structure.\nWe empirically show that our proposed predictive independence test outperform\nor are on par to current practice, and the derived graphical model structure\nlearning algorithms asymptotically recover the 'true' graph. This paper, and\nthe 'pcit' package accompanying it, thus provide powerful, scalable,\ngeneralizable, and easy-to-use methods for multivariate and conditional\nindependence testing, as well as for graphical model structure learning.\n",
"title": "Predictive Independence Testing, Predictive Conditional Independence Testing, and Predictive Graphical Modelling"
}
| null | null | null | null | true | null |
15964
| null |
Default
| null | null |
null |
{
"abstract": " Despite their overwhelming capacity to overfit, deep learning architectures\ntend to generalize relatively well to unseen data, allowing them to be deployed\nin practice. However, explaining why this is the case is still an open area of\nresearch. One standing hypothesis that is gaining popularity, e.g. Hochreiter &\nSchmidhuber (1997); Keskar et al. (2017), is that the flatness of minima of the\nloss function found by stochastic gradient based methods results in good\ngeneralization. This paper argues that most notions of flatness are problematic\nfor deep models and can not be directly applied to explain generalization.\nSpecifically, when focusing on deep networks with rectifier units, we can\nexploit the particular geometry of parameter space induced by the inherent\nsymmetries that these architectures exhibit to build equivalent models\ncorresponding to arbitrarily sharper minima. Furthermore, if we allow to\nreparametrize a function, the geometry of its parameters can change drastically\nwithout affecting its generalization properties.\n",
"title": "Sharp Minima Can Generalize For Deep Nets"
}
| null | null | null | null | true | null |
15965
| null |
Default
| null | null |
null |
{
"abstract": " The J-integral is recognized as a fundamental parameter in fracture mechanics\nthat characterizes the inherent resistance of materials to crack growth.\nHowever, the conventional methods to calculate the J-integral, which require\nknowledge of the exact position of a crack tip and the continuum fields around\nit, are unable to precisely measure the J-integral of polymer composites at the\nnanoscale. This work aims to propose an effective calculation method based on\ncoarse-grained (CG) simulations for predicting the J-integral of carbon\nnanotube (CNT)/polymer composites. In the proposed approach, the J-integral is\ndetermined from the load displacement curve of a single specimen. The\ndistinguishing feature of the method is the calculation of J-integral without\nneed of information about the crack tip, which makes it applicable to complex\npolymer systems. The effects of the CNT weight fraction and covalent\ncross-links between the polymer matrix and nanotubes, and polymer chains on the\nfracture behavior of the composites are studied in detail. The dependence of\nthe J-integral on the crack length and the size of representative volume\nelement (RVE) is also explored.\n",
"title": "Coarse-grained model of the J-integral of carbon nanotube reinforced polymer composites"
}
| null | null | null | null | true | null |
15966
| null |
Default
| null | null |
null |
{
"abstract": " Revealed preference theory studies the possibility of modeling an agent's\nrevealed preferences and the construction of a consistent utility function.\nHowever, modeling agent's choices over preference orderings is not always\npractical and demands strong assumptions on human rationality and\ndata-acquisition abilities. Therefore, we propose a simple generative choice\nmodel where agents are assumed to generate the choice probabilities based on\nlatent factor matrices that capture their choice evaluation across multiple\nattributes. Since the multi-attribute evaluation is typically hidden within the\nagent's psyche, we consider a signaling mechanism where agents are provided\nwith choice information through private signals, so that the agent's choices\nprovide more insight about his/her latent evaluation across multiple\nattributes. We estimate the choice model via a novel multi-stage matrix\nfactorization algorithm that minimizes the average deviation of the factor\nestimates from choice data. Simulation results are presented to validate the\nestimation performance of our proposed algorithm.\n",
"title": "On Estimating Multi-Attribute Choice Preferences using Private Signals and Matrix Factorization"
}
| null | null | null | null | true | null |
15967
| null |
Default
| null | null |
null |
{
"abstract": " Our view of the universe of genomic regions harboring various types of\ncandidate human-specific regulatory sequences (HSRS) has been markedly expanded\nin recent years. To infer the evolutionary origins of loci harboring HSRS,\nanalyses of conservations patterns of 59,732 loci in Modern Humans, Chimpanzee,\nBonobo, Gorilla, Orangutan, Gibbon, and Rhesus genomes have been performed. Two\nmajor evolutionary pathways have been identified comprising thousands of\nsequences that were either inherited from extinct common ancestors (ECAs) or\ncreated de novo in humans after human/chimpanzee split. Thousands of HSRS\nappear inherited from ECAs yet bypassed genomes of our closest evolutionary\nrelatives, presumably due to the incomplete lineage sorting and/or\nspecies-specific loss or regulatory DNA. The bypassing pattern is prominent for\nHSRS associated with development and functions of human brain. Common genomic\nloci that may contributed to speciation during evolution of Great Apes comprise\n248 insertions sites of African Great Ape-specific retrovirus PtERV1 (45.9%; p\n= 1.03E-44) intersecting regions harboring 442 HSRS, which are enriched for\nHSRS associated with human-specific (HS) changes of gene expression in cerebral\norganoids. Among non-human primates (NHP), most significant fractions of\ncandidate HSRS associated with HS expression changes in both excitatory neurons\n(347 loci; 67%) and radial glia (683 loci; 72%) are highly conserved in Gorilla\ngenome. Modern Humans acquired unique combinations of regulatory sequences\nhighly conserved in distinct species of six NHP separated by 30 million years\nof evolution. Concurrently, this unique mosaic of regulatory sequences\ninherited from ECAs was supplemented with 12,486 created de novo HSRS. These\nobservations support the model of complex continuous speciation process during\nevolution of Great Apes that is not likely to occur as an instantaneous event.\n",
"title": "Analysis of evolutionary origins of genomic loci harboring 59,732 candidate human-specific regulatory sequences identifies genetic divergence patterns during evolution of Great Apes"
}
| null | null | null | null | true | null |
15968
| null |
Default
| null | null |
null |
{
"abstract": " In this article, we discuss a verification study of an operational solar\nflare forecast in the Regional Warning Center (RWC) Japan. The RWC Japan has\nbeen issuing four-categorical deterministic solar flare forecasts for a long\ntime. In this forecast verification study, we used solar flare forecast data\naccumulated over 16 years (from 2000 to 2015). We compiled the forecast data\ntogether with solar flare data obtained with the Geostationary Operational\nEnvironmental Satellites (GOES). Using the compiled data sets, we estimated\nsome conventional scalar verification measures with 95% confidence intervals.\nWe also estimated a multi-categorical scalar verification measure. These scalar\nverification measures were compared with those obtained by the persistence\nmethod and recurrence method. As solar activity varied during the 16 years, we\nalso applied verification analyses to four subsets of forecast-observation pair\ndata with different solar activity levels. We cannot conclude definitely that\nthere are significant performance difference between the forecasts of RWC Japan\nand the persistence method, although a slightly significant difference is found\nfor some event definitions. We propose to use a scalar verification measure to\nassess the judgment skill of the operational solar flare forecast. Finally, we\npropose a verification strategy for deterministic operational solar flare\nforecasting.\n",
"title": "Verification of operational solar flare forecast: Case of Regional Warning Center Japan"
}
| null | null |
[
"Physics",
"Statistics"
] | null | true | null |
15969
| null |
Validated
| null | null |
null |
{
"abstract": " We use plasmon rulers to follow the conformational dynamics of a single\nprotein for up to 24 h at a video rate. The plasmon ruler consists of two gold\nnanospheres connected by a single protein linker. In our experiment, we follow\nthe dynamics of the molecular chaperone heat shock protein 90, which is known\nto show open and closed conformations. Our measurements confirm the previously\nknown conformational dynamics with transition times in the second to minute\ntime scale and reveals new dynamics on the time scale of minutes to hours.\nPlasmon rulers thus extend the observation bandwidth 3/4 orders of magnitude\nwith respect to single-molecule fluorescence resonance energy transfer and\nenable the study of molecular dynamics with unprecedented precision.\n",
"title": "Conformational dynamics of a single protein monitored for 24 hours at video rate"
}
| null | null | null | null | true | null |
15970
| null |
Default
| null | null |
null |
{
"abstract": " It is shown that the total set of equations, which determines the dynamics of\nthe domain bounds (DB) in a weak ferromagnet, has the same type of specific\nsolution as the well-known Walker's solution for ferromagnets. We calculated\nthe functional dependence of the velocity of the DB on the magnetic field,\nwhich is described by the obtained solution. This function has a maximum at a\nfinite field and a section of the negative differential mobility of the DB.\nAccording to the calculation, the maximum velocity $ c \\approx 2 \\times 10^6$\ncm/sec in YFeO$_3$ is reached at $H_m \\approx 4 \\times 10^3$ Oe.\n",
"title": "Dynamics of domain walls in weak ferromagnets"
}
| null | null | null | null | true | null |
15971
| null |
Default
| null | null |
null |
{
"abstract": " Short circuit ratio (SCR) is widely applied to analyze the strength of AC\nsystem and the small signal stability for single power elec-tronic based\ndevices infeed systems (SPEISs). However, there still lacking the theory of\nshort circuit ratio applicable for multi power electronic based devices infeed\nsystems (MPEIS), as the complex coupling among multi power electronic devices\n(PEDs) leads to difficulties in stability analysis. In this regard, this paper\nfirstly proposes a concept named generalized short circuit ratio (gSCR) to\nmeasure the strength of connected AC grid in a multi-infeed system from the\nsmall signal stability point of view. Generally, the gSCR is physically and\nmathematically extended from conven-tional SCR by decomposing the multi-infeed\nsystem into n inde-pendent single infeed systems. Then the operation gSCR\n(OgSCR) is proposed based on gSCR in order to take the variation of op-eration\npoint into consideration. The participation factors and sensitivity are\nanalyzed as well. Finally, simulations are conducted to demonstrate the\nrationality and effectiveness of the defined gSCR and OgSCR.\n",
"title": "Generalized Short Circuit Ratio for Multi Power Electronic based Devices Infeed Systems: Defi-nition and Theoretical Analysis"
}
| null | null | null | null | true | null |
15972
| null |
Default
| null | null |
null |
{
"abstract": " In this work, we addressed the issue of applying a stochastic classifier and\na local, fuzzy confusion matrix under the framework of multi-label\nclassification. We proposed a novel solution to the problem of correcting label\npairwise ensembles. The main step of the correction procedure is to compute\nclassifier- specific competence and cross-competence measures, which estimates\nerror pattern of the underlying classifier. We considered two improvements of\nthe method of obtaining confusion matrices. The first one is aimed to deal with\nimbalanced labels. The other utilizes double labelled instances which are\nusually removed during the pairwise transformation. The proposed methods were\nevaluated using 29 benchmark datasets. In order to assess the efficiency of the\nintroduced models, they were compared against 1 state-of-the-art approach and\nthe correction scheme based on the original method of confusion matrix\nestimation. The comparison was performed using four different multi-label\nevaluation measures: macro and micro-averaged F1 loss, zero-one loss and\nHamming loss. Additionally, we investigated relations between classification\nquality, which is expressed in terms of different quality criteria, and\ncharacteristics of multi-label datasets such as average imbalance ratio or\nlabel density. The experimental study reveals that the correction approaches\nsignificantly outperforms the reference method only in terms of zero-one loss.\n",
"title": "A Correction Method of a Binary Classifier Applied to Multi-label Pairwise Models"
}
| null | null | null | null | true | null |
15973
| null |
Default
| null | null |
null |
{
"abstract": " Large-batch SGD is important for scaling training of deep neural networks.\nHowever, without fine-tuning hyperparameter schedules, the generalization of\nthe model may be hampered. We propose to use batch augmentation: replicating\ninstances of samples within the same batch with different data augmentations.\nBatch augmentation acts as a regularizer and an accelerator, increasing both\ngeneralization and performance scaling. We analyze the effect of batch\naugmentation on gradient variance and show that it empirically improves\nconvergence for a wide variety of deep neural networks and datasets. Our\nresults show that batch augmentation reduces the number of necessary SGD\nupdates to achieve the same accuracy as the state-of-the-art. Overall, this\nsimple yet effective method enables faster training and better generalization\nby allowing more computational resources to be used concurrently.\n",
"title": "Augment your batch: better training with larger batches"
}
| null | null | null | null | true | null |
15974
| null |
Default
| null | null |
null |
{
"abstract": " Recent years have seen a flurry of activities in designing provably efficient\nnonconvex procedures for solving statistical estimation problems. Due to the\nhighly nonconvex nature of the empirical loss, state-of-the-art procedures\noften require proper regularization (e.g. trimming, regularized cost,\nprojection) in order to guarantee fast convergence. For vanilla procedures such\nas gradient descent, however, prior theory either recommends highly\nconservative learning rates to avoid overshooting, or completely lacks\nperformance guarantees.\nThis paper uncovers a striking phenomenon in nonconvex optimization: even in\nthe absence of explicit regularization, gradient descent enforces proper\nregularization implicitly under various statistical models. In fact, gradient\ndescent follows a trajectory staying within a basin that enjoys nice geometry,\nconsisting of points incoherent with the sampling mechanism. This \"implicit\nregularization\" feature allows gradient descent to proceed in a far more\naggressive fashion without overshooting, which in turn results in substantial\ncomputational savings. Focusing on three fundamental statistical estimation\nproblems, i.e. phase retrieval, low-rank matrix completion, and blind\ndeconvolution, we establish that gradient descent achieves near-optimal\nstatistical and computational guarantees without explicit regularization. In\nparticular, by marrying statistical modeling with generic optimization theory,\nwe develop a general recipe for analyzing the trajectories of iterative\nalgorithms via a leave-one-out perturbation argument. As a byproduct, for noisy\nmatrix completion, we demonstrate that gradient descent achieves near-optimal\nerror control --- measured entrywise and by the spectral norm --- which might\nbe of independent interest.\n",
"title": "Implicit Regularization in Nonconvex Statistical Estimation: Gradient Descent Converges Linearly for Phase Retrieval, Matrix Completion and Blind Deconvolution"
}
| null | null | null | null | true | null |
15975
| null |
Default
| null | null |
null |
{
"abstract": " New numerical solutions to the so-called selection problem for one and two\nsteadily translating bubbles in an unbounded Hele-Shaw cell are presented. Our\napproach relies on conformal mapping which, for the two-bubble problem,\ninvolves the Schottky-Klein prime function associated with an annulus. We show\nthat a countably infinite number of solutions exist for each fixed value of\ndimensionless surface tension, with the bubble shapes becoming more exotic as\nthe solution branch number increases. Our numerical results suggest that a\nsingle solution is selected in the limit that surface tension vanishes, with\nthe scaling between the bubble velocity and surface tension being different to\nthe well-studied problems for a bubble or a finger propagating in a channel\ngeometry.\n",
"title": "The effect of surface tension on steadily translating bubbles in an unbounded Hele-Shaw cell"
}
| null | null | null | null | true | null |
15976
| null |
Default
| null | null |
null |
{
"abstract": " We report on SPT-CLJ2011-5228, a giant system of arcs created by a cluster at\n$z=1.06$. The arc system is notable for the presence of a bright central image.\nThe source is a Lyman Break galaxy at $z_s=2.39$ and the mass enclosed within\nthe 14 arc second radius Einstein ring is $10^{14.2}$ solar masses. We perform\na full light profile reconstruction of the lensed images to precisely infer the\nparameters of the mass distribution. The brightness of the central image\ndemands that the central total density profile of the lens be shallow. By\nfitting the dark matter as a generalized Navarro-Frenk-White profile---with a\nfree parameter for the inner density slope---we find that the break radius is\n$270^{+48}_{-76}$ kpc, and that the inner density falls with radius to the\npower $-0.38\\pm0.04$ at 68 percent confidence. Such a shallow profile is in\nstrong tension with our understanding of relaxed cold dark matter halos; dark\nmatter only simulations predict the inner density should fall as $r^{-1}$. The\ntension can be alleviated if this cluster is in fact a merger; a two halo model\ncan also reconstruct the data, with both clumps (density going as $r^{-0.8}$\nand $r^{-1.0}$) much more consistent with predictions from dark matter only\nsimulations. At the resolution of our Dark Energy Survey imaging, we are unable\nto choose between these two models, but we make predictions for forthcoming\nHubble Space Telescope imaging that will decisively distinguish between them.\n",
"title": "Core or cusps: The central dark matter profile of a redshift one strong lensing cluster with a bright central image"
}
| null | null | null | null | true | null |
15977
| null |
Default
| null | null |
null |
{
"abstract": " One of the most interesting features of Bayesian optimization for direct\npolicy search is that it can leverage priors (e.g., from simulation or from\nprevious tasks) to accelerate learning on a robot. In this paper, we are\ninterested in situations for which several priors exist but we do not know in\nadvance which one fits best the current situation. We tackle this problem by\nintroducing a novel acquisition function, called Most Likely Expected\nImprovement (MLEI), that combines the likelihood of the priors and the expected\nimprovement. We evaluate this new acquisition function on a transfer learning\ntask for a 5-DOF planar arm and on a possibly damaged, 6-legged robot that has\nto learn to walk on flat ground and on stairs, with priors corresponding to\ndifferent stairs and different kinds of damages. Our results show that MLEI\neffectively identifies and exploits the priors, even when there is no obvious\nmatch between the current situations and the priors.\n",
"title": "Bayesian Optimization with Automatic Prior Selection for Data-Efficient Direct Policy Search"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
15978
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper we prove the existence of a nonnegative ground state solution\nto the following class of coupled systems involving Schrödinger equations\nwith square root of the Laplacian\n$$\n\\left\\{\n\\begin{array}{lr}\n(-\\Delta)^{1/2}u+V_{1}(x)u=f_{1}(u)+\\lambda(x)v, & x\\in\\mathbb{R},\n(-\\Delta)^{1/2}v+V_{2}(x)v=f_{2}(v)+\\lambda(x)u, & x\\in\\mathbb{R},\n\\end{array}\n\\right.\n$$\nwhere the nonlinearities $f_{1}(s)$ and $f_{2}(s)$ have exponential critical\ngrowth of the Trudinger-Moser type, the potentials $V_{1}(x)$ and $V_{2}(x)$\nare nonnegative and periodic. Moreover, we assume that there exists $\\delta\\in\n(0,1)$ such that $\\lambda(x)\\leq\\delta\\sqrt{V_{1}(x)V_{2}(x)}$. We are also\nconcerned with the existence of ground states when the potentials are\nasymptotically periodic. Our approach is variational and based on minimization\ntechnique over the Nehari manifold.\n",
"title": "Coupled elliptic systems involving the square root of the Laplacian and Trudinger-Moser critical growth"
}
| null | null | null | null | true | null |
15979
| null |
Default
| null | null |
null |
{
"abstract": " The paper aims to apply the complex octonion to explore the influence of the\nenergy gradient on the Eotvos experiment, impacting the gravitational mass in\nthe ultra-strong magnetic fields. Until now the Eotvos experiment has never\nbeen validated under the ultra-strong magnetic field. It is aggravating the\nexisting serious qualms about the Eotvos experiment. According to the\nelectromagnetic and gravitational theory described with the complex octonions,\nthe ultra-strong magnetic field must result in a tiny variation of the\ngravitational mass. The magnetic field with the gradient distribution will\ngenerate the energy gradient. These influencing factors will exert an influence\non the state of equilibrium in the Eotvos experiment. That is, the\ngravitational mass will depart from the inertial mass to a certain extent, in\nthe ultra-strong magnetic fields. Only under exceptional circumstances,\nespecially in the case of the weak field strength, the gravitational mass may\nbe equal to the inertial mass approximately. The paper appeals intensely to\nvalidate the Eotvos experiment in the ultra-strong electromagnetic strengths.\nIt is predicted that the physical property of gravitational mass will be\ndistinct from that of inertial mass.\n",
"title": "Gravitational mass and energy gradient in the ultra-strong magnetic fields"
}
| null | null | null | null | true | null |
15980
| null |
Default
| null | null |
null |
{
"abstract": " Knowledge graphs are a versatile framework to encode richly structured data\nrelationships, but it can be challenging to combine these graphs with\nunstructured data. Methods for retrofitting pre-trained entity representations\nto the structure of a knowledge graph typically assume that entities are\nembedded in a connected space and that relations imply similarity. However,\nuseful knowledge graphs often contain diverse entities and relations (with\npotentially disjoint underlying corpora) which do not accord with these\nassumptions. To overcome these limitations, we present Functional Retrofitting,\na framework that generalizes current retrofitting methods by explicitly\nmodeling pairwise relations. Our framework can directly incorporate a variety\nof pairwise penalty functions previously developed for knowledge graph\ncompletion. Further, it allows users to encode, learn, and extract information\nabout relation semantics. We present both linear and neural instantiations of\nthe framework. Functional Retrofitting significantly outperforms existing\nretrofitting methods on complex knowledge graphs and loses no accuracy on\nsimpler graphs (in which relations do imply similarity). Finally, we\ndemonstrate the utility of the framework by predicting new drug--disease\ntreatment pairs in a large, complex health knowledge graph.\n",
"title": "Retrofitting Distributional Embeddings to Knowledge Graphs with Functional Relations"
}
| null | null | null | null | true | null |
15981
| null |
Default
| null | null |
null |
{
"abstract": " With the emerging of smart grid techniques, cyber attackers may be able to\ngain access to critical energy infrastructure data and strategic market\nparticipants may be able to identify offer prices of their rivals. This paper\ndiscusses a privacy-preserving economic dispatch approach in competitive\nelectricity market, in which individual generation companies (GENCOs) and load\nserving entities (LSEs) can mask their actual bidding information and physical\ndata by multiplying with random numbers before submitting to Independent System\nOperators (ISOs) and Regional Transmission Owners (RTOs). This would avoid\npotential information leakage of critical energy infrastructure and financial\ndata of market participants. The optimal solution to the original ED problem,\nincluding optimal dispatches of generators and loads and locational marginal\nprices (LMPs), can be retrieved from the optimal solution of the proposed\nprivacy-preserving ED approach. Numerical case studies show the effectiveness\nof the proposed approach for protecting private information of individual\nmarket participants while guaranteeing the same optimal ED solution.\nComputation and communication costs of the proposed privacy-preserving ED\napproach and the original ED are also compared in case studies.\n",
"title": "Privacy-Preserving Economic Dispatch in Competitive Electricity Market"
}
| null | null | null | null | true | null |
15982
| null |
Default
| null | null |
null |
{
"abstract": " Thanks to multi-spacecraft mission, it has recently been possible to directly\nestimate the current density in space plasmas, by using magnetic field time\nseries from four satellites flying in a quasi perfect tetrahedron\nconfiguration. The technique developed, commonly called 'curlometer' permits a\ngood estimation of the current density when the magnetic field time series vary\nlinearly in space. This approximation is generally valid for small spacecraft\nseparation. The recent space missions Cluster and Magnetospheric Multiscale\n(MMS) have provided high resolution measurements with inter-spacecraft\nseparation up to 100 km and 10 km, respectively. The former scale corresponds\nto the proton gyroradius/ion skin depth in 'typical' solar wind conditions,\nwhile the latter to sub-proton scale. However, some works have highlighted an\nunderestimation of the current density via the curlometer technique with\nrespect to the current computed directly from the velocity distribution\nfunctions, measured at sub-proton scales resolution with MMS. In this paper we\nexplore the limit of the curlometer technique studying synthetic data sets\nassociated to a cluster of four artificial satellites allowed to fly in a\nstatic turbulent field, spanning a wide range of relative separation. This\nstudy tries to address the relative importance of measuring plasma moments at\nvery high resolution from a single spacecraft with respect to the\nmulti-spacecraft missions in the current density evaluation.\n",
"title": "On the estimation of the current density in space plasmas: multi versus single-point techniques"
}
| null | null |
[
"Physics"
] | null | true | null |
15983
| null |
Validated
| null | null |
null |
{
"abstract": " For a field $k$, we prove that the $i$th homology of the groups $GL_n(k)$,\n$SL_n(k)$, $Sp_{2n}(k)$, $SO_{n,n}(k)$, and $SO_{n,n+1}(k)$ with coefficients\nin their Steinberg representations vanish for $n \\geq 2i+2$.\n",
"title": "Homological vanishing for the Steinberg representation"
}
| null | null | null | null | true | null |
15984
| null |
Default
| null | null |
null |
{
"abstract": " High-signal to noise observations of the Ly$\\alpha$ forest transmissivity in\nthe z = 7.085 QSO ULAS J1120+0641 show seven narrow transmission spikes\nfollowed by a long 240 cMpc/h trough. Here we use radiative transfer\nsimulations of cosmic reionization previously calibrated to match a wider range\nof Ly$\\alpha$ forest data to show that the occurrence of seven transmission\nspikes in the narrow redshift range z = 5.85 - 6.1 is very sensitive to the\nexact timing of reionization. Occurrence of the spikes requires the most under\ndense regions of the IGM to be already fully ionised. The rapid onset of a long\ntrough at z = 6.12 requires a strong decrease of the photo-ionisation rate at\nz$\\sim$6.1 in this line-of-sight, consistent with the end of percolation at\nthis redshift. The narrow range of reionisation histories that we previously\nfound to be consistent with a wider range of Ly$\\alpha$ forest data have a\nreasonable probability of showing seven spikes and the mock absorption spectra\nprovide an excellent match to the spikes and the trough in the observed\nspectrum of ULAS J1120+0641. Despite the large overall opacity of Ly$\\alpha$ at\nz > 5.8, larger samples of high signal-to-noise observations of rare\ntransmission spikes should therefore provide important further insights into\nthe exact timing of the percolation of HII bubbles at the tail-end of\nreionization\n",
"title": "A tale of seven narrow spikes and a long trough: constraining the timing of the percolation of HII bubbles at the tail-end of reionization with ULAS J1120+0641"
}
| null | null |
[
"Physics"
] | null | true | null |
15985
| null |
Validated
| null | null |
null |
{
"abstract": " The intermediate-valence compound SmB6 is a well-known Kondo insulator, in\nwhich hybridization of itinerant 5d electrons with localized 4f electrons leads\nto a transition from metallic to insulating behavior at low temperatures.\nRecent studies suggest that SmB6 is a topological insulator, with topological\nmetallic surface states emerging from a fully insulating hybridized bulk band\nstructure. Here we locally probe the bulk magnetic properties of pure and 0.5 %\nFe-doped SmB6 by muon spin rotation/relaxation methods. Below 6 K the Fe\nimpurity induces simultaneous changes in the bulk local magnetism and the\nelectrical conductivity. In the low-temperature insulating bulk state we\nobserve a temperature-independent dynamic relaxation rate indicative of\nlow-lying magnetic excitations driven primarily by quantum fluctuations.\n",
"title": "Quantum spin fluctuations in the bulk insulating state of pure and Fe-doped SmB6"
}
| null | null | null | null | true | null |
15986
| null |
Default
| null | null |
null |
{
"abstract": " The dramatic increase in data and connectivity demand, in addition to\nheterogeneous device capabilities, poses a challenge for future wireless\nnetworks. One of the promising solutions is Device-to-Device (D2D) networking.\nD2D networking, advocating the idea of connecting two or more devices directly\nwithout traversing the core network, is promising to address the increasing\ndata and connectivity demand. In this paper, we consider D2D networks, where\ndevices with heterogeneous capabilities including computing power, energy\nlimitations, and incentives participate in D2D activities heterogeneously. We\ndevelop (i) a device-aware routing and scheduling algorithm (DARS) by taking\ninto account device capabilities, and (ii) a multi-hop D2D testbed using\nAndroid-based smartphones and tablets by exploiting Wi-Fi Direct and legacy\nWi-Fi connections. We show that DARS significantly improves throughput in our\ntestbed as compared to state-of-the-art.\n",
"title": "Device-Aware Routing and Scheduling in Multi-Hop Device-to-Device Networks"
}
| null | null | null | null | true | null |
15987
| null |
Default
| null | null |
null |
{
"abstract": " Techniques known as Nonlinear Set Membership prediction, Lipschitz\nInterpolation or Kinky Inference are approaches to machine learning that\nutilise presupposed Lipschitz properties to compute inferences over unobserved\nfunction values. Provided a bound on the true best Lipschitz constant of the\ntarget function is known a priori they offer convergence guarantees as well as\nbounds around the predictions. Considering a more general setting that builds\non Hoelder continuity relative to pseudo-metrics, we propose an online method\nfor estimating the Hoelder constant online from function value observations\nthat possibly are corrupted by bounded observational errors. Utilising this to\ncompute adaptive parameters within a kinky inference rule gives rise to a\nnonparametric machine learning method, for which we establish strong universal\napproximation guarantees. That is, we show that our prediction rule can learn\nany continuous function in the limit of increasingly dense data to within a\nworst-case error bound that depends on the level of observational uncertainty.\nWe apply our method in the context of nonparametric model-reference adaptive\ncontrol (MRAC). Across a range of simulated aircraft roll-dynamics and\nperformance metrics our approach outperforms recently proposed alternatives\nthat were based on Gaussian processes and RBF-neural networks. For\ndiscrete-time systems, we provide guarantees on the tracking success of our\nlearning-based controllers both for the batch and the online learning setting.\n",
"title": "Lazily Adapted Constant Kinky Inference for Nonparametric Regression and Model-Reference Adaptive Control"
}
| null | null | null | null | true | null |
15988
| null |
Default
| null | null |
null |
{
"abstract": " Despite enormous progress in object detection and classification, the problem\nof incorporating expected contextual relationships among object instances into\nmodern recognition systems remains a key challenge. In this work we propose\nInformation Pursuit, a Bayesian framework for scene parsing that combines prior\nmodels for the geometry of the scene and the spatial arrangement of objects\ninstances with a data model for the output of high-level image classifiers\ntrained to answer specific questions about the scene. In the proposed\nframework, the scene interpretation is progressively refined as evidence\naccumulates from the answers to a sequence of questions. At each step, we\nchoose the question to maximize the mutual information between the new answer\nand the full interpretation given the current evidence obtained from previous\ninquiries. We also propose a method for learning the parameters of the model\nfrom synthesized, annotated scenes obtained by top-down sampling from an\neasy-to-learn generative scene model. Finally, we introduce a database of\nannotated indoor scenes of dining room tables, which we use to evaluate the\nproposed approach.\n",
"title": "Information Pursuit: A Bayesian Framework for Sequential Scene Parsing"
}
| null | null | null | null | true | null |
15989
| null |
Default
| null | null |
null |
{
"abstract": " This paper considers a network of sensors without fusion center that may be\ndifficult to set up in applications involving sensors embedded on autonomous\ndrones or robots. In this context, this paper considers that the sensors must\nperform a given clustering task in a fully decentralized setup. Standard\nclustering algorithms usually need to know the number of clusters and are very\nsensitive to initialization, which makes them difficult to use in a fully\ndecentralized setup. In this respect, this paper proposes a decentralized\nmodel-based clustering algorithm that overcomes these issues. The proposed\nalgorithm is based on a novel theoretical framework that relies on hypothesis\ntesting and robust M-estimation. More particularly, the problem of deciding\nwhether two data belong to the same cluster can be optimally solved via Wald's\nhypothesis test on the mean of a Gaussian random vector. The p-value of this\ntest makes it possible to define a new type of score function, particularly\nsuitable for devising an M-estimation of the centroids. The resulting\ndecentralized algorithm efficiently performs clustering without prior knowledge\nof the number of clusters. It also turns out to be less sensitive to\ninitialization than the already existing clustering algorithms, which makes it\nappropriate for use in a network of sensors without fusion center.\n",
"title": "Decentralized Clustering based on Robust Estimation and Hypothesis Testing"
}
| null | null | null | null | true | null |
15990
| null |
Default
| null | null |
null |
{
"abstract": " Recently, an open geometry Fourier modal method based on a new combination of\nan open boundary condition and a non-uniform $k$-space discretization was\nintroduced for rotationally symmetric structures providing a more efficient\napproach for modeling nanowires and micropillar cavities [J. Opt. Soc. Am. A\n33, 1298 (2016)]. Here, we generalize the approach to three-dimensional (3D)\nCartesian coordinates allowing for the modeling of rectangular geometries in\nopen space. The open boundary condition is a consequence of having an infinite\ncomputational domain described using basis functions that expand the whole\nspace. The strength of the method lies in discretizing the Fourier integrals\nusing a non-uniform circular \"dartboard\" sampling of the Fourier $k$ space. We\nshow that our sampling technique leads to a more accurate description of the\ncontinuum of the radiation modes that leak out from the structure. We also\ncompare our approach to conventional discretization with direct and inverse\nfactorization rules commonly used in established Fourier modal methods. We\napply our method to a variety of optical waveguide structures and demonstrate\nthat the method leads to a significantly improved convergence enabling more\naccurate and efficient modeling of open 3D nanophotonic structures.\n",
"title": "Modeling open nanophotonic systems using the Fourier modal method: Generalization to 3D Cartesian coordinates"
}
| null | null |
[
"Physics"
] | null | true | null |
15991
| null |
Validated
| null | null |
null |
{
"abstract": " We apply Lieb-Robinson bounds for multi-commutators we recently derived to\nstudy the (possibly non-linear) response of interacting fermions at thermal\nequilibrium to perturbations of the external electromagnetic field. This\nanalysis leads to an extension of the results for quasi-free fermions of\n\\cite{OhmI,OhmII} to fermion systems on the lattice with short-range\ninteractions. More precisely, we investigate entropy production and charge\ntransport properties of non-autonomous $C^{\\ast }$-dynamical systems associated\nwith interacting lattice fermions within bounded static potentials and in\npresence of an electric field that is time- and space-dependent. We verify the\n1st law of thermodynamics for the heat production of the system under\nconsideration. In linear response theory, the latter is related with Ohm and\nJoule's laws. These laws are proven here to hold at the microscopic scale,\nuniformly with respect to the size of the (microscopic) region where the\nelectric field is applied. An important outcome is the extension of the notion\nof conductivity measures to interacting fermions.\n",
"title": "Microscopic Conductivity of Lattice Fermions at Equilibrium - Part II: Interacting Particles"
}
| null | null | null | null | true | null |
15992
| null |
Default
| null | null |
null |
{
"abstract": " Cloud users have little visibility into the performance characteristics and\nutilization of the physical machines underpinning the virtualized cloud\nresources they use. This uncertainty forces users and researchers to reverse\nengineer the inner workings of cloud systems in order to understand and\noptimize the conditions their applications operate. At Massachusetts Open Cloud\n(MOC), as a public cloud operator, we'd like to expose the utilization of our\nphysical infrastructure to stop this wasteful effort. Mindful that such\nexposure can be used maliciously for gaining insight into other users\nworkloads, in this position paper we argue for the need for an approach that\nbalances openness of the cloud overall with privacy for each tenant inside of\nit. We believe that this approach can be instantiated via a novel combination\nof several security and privacy technologies. We discuss the potential\nbenefits, implications of transparency for cloud systems and users, and\ntechnical challenges/possibilities.\n",
"title": "Revealing the Unseen: How to Expose Cloud Usage While Protecting User Privacy"
}
| null | null |
[
"Computer Science"
] | null | true | null |
15993
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper a new hp-adaptive strategy for elliptic problems based on\nrefinement history is proposed, which chooses h-, p- or hp-refinement on\nindividual elements according to a posteriori error estimate, as well as\nsmoothness estimate of the solution obtained by comparing the actual and\nexpected error reduction rate. Numerical experiments show that exponential\nconvergence can be achieved with this strategy.\n",
"title": "An hp-adaptive strategy for elliptic problems"
}
| null | null | null | null | true | null |
15994
| null |
Default
| null | null |
null |
{
"abstract": " Recent space missions have provided information on the physical and chemical\nproperties of interstellar grains such as the ratio $\\beta$ of radiation\npressure to gravity acting on the grains in addition to the composition,\nstructure, and size distribution of the grains. Numerical simulation on the\ntrajectories of interstellar grains captured by Stardust and returned to Earth\nconstrained the $\\beta$ ratio for the Stardust samples of interstellar origin.\nHowever, recent accurate calculations of radiation pressure cross sections for\nmodel dust grains have given conflicting stories in the $\\beta$ ratio of\ninterstellar grains. The $\\beta$ ratio for model dust grains of so-called\n\"astronomical silicate\" in the femto-kilogram range lies below unity, in\nconflict with $\\beta \\sim 1$ for the Stardust interstellar grains. Here, I\ntackle this conundrum by re-evaluating the $\\beta$ ratio of interstellar grains\non the assumption that the grains are aggregated particles grown by coagulation\nand composed of amorphous MgSiO$_{3}$ with the inclusion of metallic iron. My\nmodel is entirely consistent with the depletion and the correlation of major\nrock-forming elements in the Local Interstellar Cloud surrounding the Sun and\nthe mineralogical identification of interstellar grains in the Stardust and\nCassini missions. I find that my model dust particles fulfill the constraints\non the $\\beta$ ratio derived from not only the Stardust mission but also the\nUlysses and Cassini missions. My results suggest that iron is not incorporated\ninto silicates but exists as metal, contrary to the majority of interstellar\ndust models available to date.\n",
"title": "High Radiation Pressure on Interstellar Dust Computed by Light-Scattering Simulation on Fluffy Agglomerates of Magnesium-silicate Grains with Metallic-iron Inclusions"
}
| null | null | null | null | true | null |
15995
| null |
Default
| null | null |
null |
{
"abstract": " Titanium dioxide (TiO2) is a wide band gap semiconducting material which is\npromising for photocatalysis. Here we present first-principles calculations to\nstudy the pressure dependence of structural and electronic properties of two\nTiO2 phases: the cotunnite-type and the Fe2P-type structure. The band gaps are\ncalculated using density functional theory (DFT) with the generalized gradient\napproximation (GGA), as well as the many-body perturbation theory with the GW\napproximation. The band gaps of both phases are found to be unexpectedly robust\nacross a broad range pressures. The corresponding pressure coefficients are\nsignificantly smaller than that of diamond and silicon carbide (SiC), whose\npressure coefficient is the smallest value ever measured by experiment. The\nrobustness originates from the synchronous change of valence band maximum (VBM)\nand conduction band minimum (CBM) with nearly identical rates of changes. A\nstep-like jump of band gaps around the phase transition pressure point is\nexpected and understood in light of the difference in crystal structures.\n",
"title": "Unexpected Robustness of the Band Gaps of TiO2 under High Pressures"
}
| null | null | null | null | true | null |
15996
| null |
Default
| null | null |
null |
{
"abstract": " Generative Adversarial Nets (GANs) represent an important milestone for\neffective generative models, which has inspired numerous variants seemingly\ndifferent from each other. One of the main contributions of this paper is to\nreveal a unified geometric structure in GAN and its variants. Specifically, we\nshow that the adversarial generative model training can be decomposed into\nthree geometric steps: separating hyperplane search, discriminator parameter\nupdate away from the separating hyperplane, and the generator update along the\nnormal vector direction of the separating hyperplane. This geometric intuition\nreveals the limitations of the existing approaches and leads us to propose a\nnew formulation called geometric GAN using SVM separating hyperplane that\nmaximizes the margin. Our theoretical analysis shows that the geometric GAN\nconverges to a Nash equilibrium between the discriminator and generator. In\naddition, extensive numerical results show that the superior performance of\ngeometric GAN.\n",
"title": "Geometric GAN"
}
| null | null | null | null | true | null |
15997
| null |
Default
| null | null |
null |
{
"abstract": " At present, the cloud storage used in searchable symmetric encryption schemes\n(SSE) is provided in a private way, which cannot be seen as a true cloud.\nMoreover, the cloud server is thought to be credible, because it always returns\nthe search result to the user, even they are not correct. In order to really\nresist this malicious adversary and accelerate the usage of the data, it is\nnecessary to store the data on a public chain, which can be seen as a\ndecentralized system. As the increasing amount of the data, the search problem\nbecomes more and more intractable, because there does not exist any effective\nsolution at present.\nIn this paper, we begin by pointing out the importance of storing the data in\na public chain. We then innovatively construct a model of SSE using\nblockchain(SSE-using-BC) and give its security definition to ensure the privacy\nof the data and improve the search efficiency. According to the size of data,\nwe consider two different cases and propose two corresponding schemes. Lastly,\nthe security and performance analyses show that our scheme is feasible and\nsecure.\n",
"title": "A Searchable Symmetric Encryption Scheme using BlockChain"
}
| null | null | null | null | true | null |
15998
| null |
Default
| null | null |
null |
{
"abstract": " The control of the electron spin by external means is a key issue for\nspintronic devices. Using spin- and angle-resolved photoemission spectroscopy\n(SARPES) with three-dimensional spin detection, we demonstrate operando\nelectrostatic spin manipulation in ferroelectric GeTe and multiferroic\nGe1-xMnxTe. We not only demonstrate for the first time electrostatic spin\nmanipulation in Rashba semiconductors due to ferroelectric polarization\nreversal, but are also able to follow the switching pathway in detail, and show\na gain of the Rashba-splitting strength under external fields. In multiferroic\nGe1-xMnxTe operando SARPES reveals switching of the perpendicular spin\ncomponent due to electric field induced magnetization reversal. This provides\nfirm evidence of effective multiferroic coupling which opens up magnetoelectric\nfunctionality with a multitude of spin-switching paths in which the magnetic\nand electric order parameters are coupled through ferroelastic relaxation\npaths. This work thus provides a new type of magnetoelectric switching\nentangled with Rashba-Zeeman splitting in a multiferroic system.\n",
"title": "Operando imaging of all-electric spin texture manipulation in ferroelectric and multiferroic Rashba semiconductors"
}
| null | null | null | null | true | null |
15999
| null |
Default
| null | null |
null |
{
"abstract": " We report on the observation of phase space modulations in the correlated\nelectron emission after strong field double ionization of helium using laser\npulses with a wavelength of 394~nm and an intensity of $3\\cdot10^{14}$W/cm$^2$.\nThose modulations are identified as direct results of quantum mechanical\nselection rules predicted by many theoretical calculations. They only occur for\nan odd number of absorbed photons. By that we attribute this effect to the\nparity of the continuum wave function.\n",
"title": "Experimental Evidence for Selection Rules in Multiphoton Double Ionization of Helium"
}
| null | null | null | null | true | null |
16000
| null |
Default
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.