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 layered transition metal dichalcogenides (LTMDCs) that display both charge\ndensity waves (CDWs) and superconductivity, the superconducting state generally\nemerges directly on suppression of the CDW state. Here, however, we report a\ndifferent observation for pressurized TaTe2, a non-superconducting CDW-bearing\nLTMDC at ambient pressure. We find that a superconducting state does not occur\nin TaTe2 after the full suppression of its CDW state, which we observe at about\n3 GPa, but, rather, a non-superconducting semimetal state is observed. At a\nhigher pressure, ~21 GPa, where both the semimetal state and the corresponding\npositive magnetoresistance effect are destroyed, superconductivity finally\nemerges and remains present up to ~50 GPa, the high pressure limit of our\nmeasurements. Our pressure-temperature phase diagram for TaTe2 demonstrates\nthat the CDW and the superconducting phases in TaTe2 do not directly transform\none to the other, but rather are separated by a semimetal state, - the first\nexperimental case where the CDW and superconducting states are separated by an\nintermediate phase in LTMDC systems.\n",
"title": "Separation of the charge density wave and superconducting states by an intermediate semimetal phase in pressurized TaTe2"
}
| null | null | null | null | true | null |
18001
| null |
Default
| null | null |
null |
{
"abstract": " We consider the problem of zero distribution of the first kind\nHermite--Padé polynomials associated with a vector function $\\vec f = (f_1,\n\\dots, f_s)$ whose components $f_k$ are functions with a finite number of\nbranch points in plane. We assume that branch sets of component functions are\nwell enough separated (which constitute the Angelesco case). Under this\ncondition we prove a theorem on limit zero distribution for such polynomials.\nThe limit measures are defined in terms of a known vector equilibrium problem.\nProof of the theorem is based on the methods developed by H.~Stahl,\nA.~A.~Gonchar and the author. These methods obtained some further\ngeneralization in the paper in application to systems of polynomials defined by\nsystems of complex orthogonality relations.\nTogether with the characterization of the limit zero distributions of\nHermite--Padé polynomials by a vector equilibrium problem we consider an\nalternative characterization using a Riemann surface $\\mathcal R(\\vec f)$\nassociated with $\\vec f$. In this terms we present a more general (without\nAngelesco condition) conjecture on the zero distribution of Hermite--Padé\npolynomials.\nBibliography: 72 items.\n",
"title": "Zero distribution for Angelesco Hermite--Padé polynomials"
}
| null | null | null | null | true | null |
18002
| null |
Default
| null | null |
null |
{
"abstract": " We have discovered two novel types of planar defects that appear in\nheteroepitaxial YBa$_2$Cu$_3$O$_{7-\\delta}$ (YBCO123) thin films, grown by\npulsed-laser deposition (PLD) either with or without a\nLa$_{2/3}$Ca$_{1/3}$MnO$_3$ (LCMO) overlayer, using the combination of\nhigh-angle annular dark-field scanning transmission electron microscopy\n(HAADF-STEM) imaging and electron energy loss spectroscopy (EELS) mapping for\nunambiguous identification. These planar lattice defects are based on the\nintergrowth of either a BaO plane between two CuO chains or multiple Y-O layers\nbetween two CuO$_2$ planes, resulting in non-stoichiometric layer sequences\nthat could directly impact the high-$T_c$ superconductivity.\n",
"title": "Atomic-scale identification of novel planar defect phases in heteroepitaxial YBa$_2$Cu$_3$O$_{7-δ}$ thin films"
}
| null | null | null | null | true | null |
18003
| null |
Default
| null | null |
null |
{
"abstract": " Our manuscript investigates a self-consistent solution of the statistical\natom model proposed by Berthold-Georg Englert and Julian Schwinger (the ES\nmodel) and benchmarks it against atomic Kohn-Sham and two orbital-free models\nof the Thomas-Fermi-Dirac (TFD)-$\\lambda$vW family. Results show that the ES\nmodel generally offers the same accuracy as the well-known TFD-$\\frac{1}{5}$vW\nmodel; however, the ES model corrects the failure in Pauli potential\nnear-nucleus region. We also point to the inability of describing low-$Z$ atoms\nas the foremost concern in improving the present model.\n",
"title": "Self-consistent assessment of Englert-Schwinger model on atomic properties"
}
| null | null | null | null | true | null |
18004
| null |
Default
| null | null |
null |
{
"abstract": " We examine in this article the pricing of target volatility options in the\nlognormal fractional SABR model. A decomposition formula by Ito's calculus\nyields a theoretical replicating strategy for the target volatility option,\nassuming the accessibilities of all variance swaps and swaptions. The same\nformula also suggests an approximation formula for the price of target\nvolatility option in small time by the technique of freezing the coefficient.\nAlternatively, we also derive closed formed expressions for a small volatility\nof volatility expansion of the price of target volatility option. Numerical\nexperiments show accuracy of the approximations in a reasonably wide range of\nparameters.\n",
"title": "Target volatility option pricing in lognormal fractional SABR model"
}
| null | null | null | null | true | null |
18005
| null |
Default
| null | null |
null |
{
"abstract": " We revisit the Blind Deconvolution problem with a focus on understanding its\nrobustness and convergence properties. Provable robustness to noise and other\nperturbations is receiving recent interest in vision, from obtaining immunity\nto adversarial attacks to assessing and describing failure modes of algorithms\nin mission critical applications. Further, many blind deconvolution methods\nbased on deep architectures internally make use of or optimize the basic\nformulation, so a clearer understanding of how this sub-module behaves, when it\ncan be solved, and what noise injection it can tolerate is a first order\nrequirement. We derive new insights into the theoretical underpinnings of blind\ndeconvolution. The algorithm that emerges has nice convergence guarantees and\nis provably robust in a sense we formalize in the paper. Interestingly, these\ntechnical results play out very well in practice, where on standard datasets\nour algorithm yields results competitive with or superior to the state of the\nart. Keywords: blind deconvolution, robust continuous optimization\n",
"title": "Robust Blind Deconvolution via Mirror Descent"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
18006
| null |
Validated
| null | null |
null |
{
"abstract": " We present a unified perspective on symmetry protected topological (SPT)\nphases in one dimension and address the open question of what characterizes\ntheir phase transitions. In the first part of this work we use symmetry as a\nguide to map various well-known fermionic and spin SPTs to a Kitaev chain with\ncoupling of range $\\alpha \\in \\mathbb Z$. This unified picture uncovers new\nproperties of old models --such as how the cluster state is the fixed point\nlimit of the Affleck-Kennedy-Lieb-Tasaki state in disguise-- and elucidates the\nconnection between fermionic and bosonic phases --with the Hubbard chain\ninterpolating between four Kitaev chains and a spin chain in the Haldane phase.\nIn the second part, we study the topological phase transitions between these\nmodels in the presence of interactions. This leads us to conjecture that the\ncritical point between any SPT with $d$-dimensional edge modes and the trivial\nphase has a central charge $c \\geq \\log_2 d$. We analytically verify this for\nmany known transitions. This agrees with the intuitive notion that the phase\ntransition is described by a delocalized edge mode, and that the central charge\nof a conformal field theory is a measure of the gapless degrees of freedom.\n",
"title": "One-Dimensional Symmetry Protected Topological Phases and their Transitions"
}
| null | null |
[
"Physics"
] | null | true | null |
18007
| null |
Validated
| null | null |
null |
{
"abstract": " This volume contains the proceedings of the Eighth International Symposium on\nGames, Automata, Logic and Formal Verification (GandALF 2017). The symposium\ntook place in Roma, Italy, from the 20th to the 22nd of September 2017. The\nGandALF symposium was established by a group of Italian computer scientists\ninterested in mathematical logic, automata theory, game theory, and their\napplications to the specification, design, and verification of complex systems.\nIts aim is to provide a forum where people from different areas, and possibly\nwith different backgrounds, can fruitfully interact. GandALF has a truly\ninternational spirit, as witnessed by the composition of the program and\nsteering committee and by the country distribution of the submitted papers.\n",
"title": "Proceedings Eighth International Symposium on Games, Automata, Logics and Formal Verification"
}
| null | null | null | null | true | null |
18008
| null |
Default
| null | null |
null |
{
"abstract": " We analyze the effect of quenched disorder on spin-1/2 quantum magnets in\nwhich magnetic frustration promotes the formation of local singlets. Our\nresults include a theory for 2d valence-bond solids subject to weak bond\nrandomness, as well as extensions to stronger disorder regimes where we make\nconnections with quantum spin liquids. We find, on various lattices, that the\ndestruction of a valence-bond solid phase by weak quenched disorder leads\ninevitably to the nucleation of topological defects carrying spin-1/2 moments.\nThis renormalizes the lattice into a strongly random spin network with\ninteresting low-energy excitations. Similarly when short-ranged valence bonds\nwould be pinned by stronger disorder, we find that this putative glass is\nunstable to defects that carry spin-1/2 magnetic moments, and whose residual\ninteractions decide the ultimate low energy fate. Motivated by these results we\nconjecture Lieb-Schultz-Mattis-like restrictions on ground states for\ndisordered magnets with spin-1/2 per statistical unit cell. These conjectures\nare supported by an argument for 1d spin chains. We apply insights from this\nstudy to the phenomenology of YbMgGaO$_4$, a recently discovered triangular\nlattice spin-1/2 insulator which was proposed to be a quantum spin liquid. We\ninstead explore a description based on the present theory. Experimental\nsignatures, including unusual specific heat, thermal conductivity, and\ndynamical structure factor, and their behavior in a magnetic field, are\npredicted from the theory, and compare favorably with existing measurements on\nYbMgGaO$_4$ and related materials.\n",
"title": "Valence Bonds in Random Quantum Magnets: Theory and Application to YbMgGaO4"
}
| null | null | null | null | true | null |
18009
| null |
Default
| null | null |
null |
{
"abstract": " Knowledge graphs are structured representations of real world facts. However,\nthey typically contain only a small subset of all possible facts. Link\nprediction is a task of inferring missing facts based on existing ones. We\npropose TuckER, a relatively simple but powerful linear model based on Tucker\ndecomposition of the binary tensor representation of knowledge graph triples.\nTuckER outperforms all previous state-of-the-art models across standard link\nprediction datasets. We prove that TuckER is a fully expressive model, deriving\nthe bound on its entity and relation embedding dimensionality for full\nexpressiveness which is several orders of magnitude smaller than the bound of\nprevious state-of-the-art models ComplEx and SimplE. We further show that\nseveral previously introduced linear models can be viewed as special cases of\nTuckER.\n",
"title": "TuckER: Tensor Factorization for Knowledge Graph Completion"
}
| null | null | null | null | true | null |
18010
| null |
Default
| null | null |
null |
{
"abstract": " We consider the multi armed bandit problem in non-stationary environments.\nBased on the Bayesian method, we propose a variant of Thompson Sampling which\ncan be used in both rested and restless bandit scenarios. Applying discounting\nto the parameters of prior distribution, we describe a way to systematically\nreduce the effect of past observations. Further, we derive the exact expression\nfor the probability of picking sub-optimal arms. By increasing the exploitative\nvalue of Bayes' samples, we also provide an optimistic version of the\nalgorithm. Extensive empirical analysis is conducted under various scenarios to\nvalidate the utility of proposed algorithms. A comparison study with various\nstate-of-the-arm algorithms is also included.\n",
"title": "Taming Non-stationary Bandits: A Bayesian Approach"
}
| null | null | null | null | true | null |
18011
| null |
Default
| null | null |
null |
{
"abstract": " We propose a new randomized coordinate descent method for a convex\noptimization template with broad applications. Our analysis relies on a novel\ncombination of four ideas applied to the primal-dual gap function: smoothing,\nacceleration, homotopy, and coordinate descent with non-uniform sampling. As a\nresult, our method features the first convergence rate guarantees among the\ncoordinate descent methods, that are the best-known under a variety of common\nstructure assumptions on the template. We provide numerical evidence to support\nthe theoretical results with a comparison to state-of-the-art algorithms.\n",
"title": "Smooth Primal-Dual Coordinate Descent Algorithms for Nonsmooth Convex Optimization"
}
| null | null | null | null | true | null |
18012
| null |
Default
| null | null |
null |
{
"abstract": " We investigate the performance of the standard Greedy algorithm for\ncardinality constrained maximization of non-submodular nondecreasing set\nfunctions. While there are strong theoretical guarantees on the performance of\nGreedy for maximizing submodular functions, there are few guarantees for\nnon-submodular ones. However, Greedy enjoys strong empirical performance for\nmany important non-submodular functions, e.g., the Bayesian A-optimality\nobjective in experimental design. We prove theoretical guarantees supporting\nthe empirical performance. Our guarantees are characterized by a combination of\nthe (generalized) curvature $\\alpha$ and the submodularity ratio $\\gamma$. In\nparticular, we prove that Greedy enjoys a tight approximation guarantee of\n$\\frac{1}{\\alpha}(1- e^{-\\gamma\\alpha})$ for cardinality constrained\nmaximization. In addition, we bound the submodularity ratio and curvature for\nseveral important real-world objectives, including the Bayesian A-optimality\nobjective, the determinantal function of a square submatrix and certain linear\nprograms with combinatorial constraints. We experimentally validate our\ntheoretical findings for both synthetic and real-world applications.\n",
"title": "Guarantees for Greedy Maximization of Non-submodular Functions with Applications"
}
| null | null |
[
"Computer Science",
"Mathematics"
] | null | true | null |
18013
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper, algebroid bundle associated to affine metrics provide an\nstructure for unification of gravity and electromagnetism and, geometrization\nof matter.\n",
"title": "Affine Metrics and Associated Algebroid Structures: Application to General Relativity"
}
| null | null |
[
"Mathematics"
] | null | true | null |
18014
| null |
Validated
| null | null |
null |
{
"abstract": " The control of complex networks is a significant challenge, especially when\nthe network topology of the system to be controlled is dynamic. Addressing this\nchallenge, here we introduce a novel approach which allows exploring the\ncontrollability of temporal networks. Studying six empirical data sets, we\nparticularly show that order correlations in the sequence of interactions can\nboth increase or decrease the time needed to achieve full controllability.\nCounter-intuitively, we find that this effect can be opposite than the effect\nof order correlations on other dynamical processes. Specifically, we show that\norder correlations that speed up a diffusion process in a given system can slow\ndown the control of the same system, and vice-versa. Building on the\nhigher-order graphical modeling framework introduced in recent works, we\nfurther demonstrate that spectral properties of higher-order network topologies\ncan be used to analytically explain this phenomenon.\n",
"title": "Controllability of temporal networks: An analysis using higher-order networks"
}
| null | null |
[
"Computer Science",
"Physics"
] | null | true | null |
18015
| null |
Validated
| null | null |
null |
{
"abstract": " Despite their fundamental role in determining material properties, detailed\nmomentum-dependent information on the strength of electron-phonon and\nphonon-phonon coupling (EPC and PPC, respectively) across the entire Brillouin\nzone (BZ) has proved difficult to obtain. Here we demonstrate that ultrafast\nelectron diffuse scattering (UEDS) directly provides such information. By\nexploiting symmetry-based selection rules and time-resolution, scattering from\ndifferent phonon branches can be distinguished even without energy resolution.\nUsing graphite as a model system, we show that UEDS patterns map the relative\nEPC and PPC strength through their profound sensitivity to photoinduced changes\nin phonon populations. We measure strong EPC to the $K$-point transverse\noptical phonon of $A_1'$ symmetry ($K-A_1'$) and along the entire longitudinal\noptical branch between $\\Gamma-K$, not only to the $\\Gamma-E_{2g}$ phonon as\npreviously emphasized. We also determine that the subsequent phonon relaxation\npathway involves three stages; decay via several identifiable channels to\ntransverse acoustic (TA) and longitudinal acoustic (LA) phonons (1-2 ps),\nintraband thermalization of the non-equilibrium TA/LA phonon populations (30-40\nps) and interband relaxation of the LA/TA modes (115 ps). Combining UEDS with\nultrafast angle-resolved photoelectron spectroscopy will yield a complete\npicture of the dynamics within and between electron and phonon subsystems,\nhelping to unravel complex phases in which the intertwined nature of these\nsystems have a strong influence on emergent properties.\n",
"title": "Mapping momentum-dependent electron-phonon coupling and non-equilibrium phonon dynamics with ultrafast electron diffuse scattering"
}
| null | null |
[
"Physics"
] | null | true | null |
18016
| null |
Validated
| null | null |
null |
{
"abstract": " Surface observations indicate that the speed of the solar meridional\ncirculation in the photosphere varies in anti-phase with the solar cycle. The\ncurrent explanation for the source of this variation is that inflows into\nactive regions alter the global surface pattern of the meridional circulation.\nWhen these localized inflows are integrated over a full hemisphere, they\ncontribute to the slow down of the axisymmetric poleward horizontal component.\nThe behavior of this large scale flow deep inside the convection zone remains\nlargely unknown. Present helioseismic techniques are not sensitive enough to\ncapture the dynamics of this weak large scale flow. Moreover, the large time of\nintegration needed to map the meridional circulation inside the convection\nzone, also masks some of the possible dynamics on shorter timescales. In this\nwork we examine the dynamics of the meridional circulation that emerges from a\n3D MHD global simulation of the solar convection zone. Our aim is to assess and\nquantify the behavior of meridional circulation deep inside the convection\nzone, where the cyclic large-scale magnetic field can reach considerable\nstrength. Our analyses indicate that the meridional circulation morphology and\namplitude are both highly influenced by the magnetic field, via the impact of\nmagnetic torques on the global angular momentum distribution. A dynamic feature\ninduced by these magnetic torques is the development of a prominent upward flow\nat mid latitudes in the lower convection zone that occurs near the equatorward\nedge of the toroidal bands and that peaks during cycle maximum. Globally, the\ndynamo-generated large-scale magnetic field drives variations in the meridional\nflow, in stark contrast to the conventional kinematic flux transport view of\nthe magnetic field being advected passively by the flow.\n",
"title": "Meridional Circulation Dynamics in a Cyclic Convective Dynamo"
}
| null | null | null | null | true | null |
18017
| null |
Default
| null | null |
null |
{
"abstract": " In many studies of environmental change of the past few centuries, 210Pb\ndating is used to obtain chronologies for sedimentary sequences. One of the\nmost commonly used approaches to estimate the ages of depths in a sequence is\nto assume a constant rate of supply (CRS) or influx of `unsupported' 210Pb from\nthe atmosphere, together with a constant or varying amount of `supported'\n210Pb. Current 210Pb dating models do not use a proper statistical framework\nand thus provide poor estimates of errors. Here we develop a new model for\n210Pb dating, where both ages and values of supported and unsupported 210Pb\nform part of the parameters. We apply our model to a case study from Canada as\nwell as to some simulated examples. Our model can extend beyond the current CRS\napproach, deal with asymmetric errors and mix 210Pb with other types of dating,\nthus obtaining more robust, realistic and statistically better defined\nestimates.\n",
"title": "Bayesian analysis of 210Pb dating"
}
| null | null |
[
"Statistics"
] | null | true | null |
18018
| null |
Validated
| null | null |
null |
{
"abstract": " The ill-posed analytic continuation problem for Green's functions or\nself-energies can be done using the Padé rational polynomial approximation.\nHowever, to extract accurate results from this approximation, high precision\ninput data of the Matsubara Green's function are needed. The calculation of the\nMatsubara Green's function generally involves a Matsubara frequency summation\nwhich cannot be evaluated analytically. Numerical summation is requisite but it\nconverges slowly with the increase of the Matsubara frequency. Here we show\nthat this slow convergence problem can be significantly improved by utilizing\nthe Padé decomposition approach to replace the Matsubara frequency summation\nby a Padé frequency summation, and high precision input data can be obtained\nto successfully perform the Padé analytic continuation.\n",
"title": "Analytic continuation with Padé decomposition"
}
| null | null | null | null | true | null |
18019
| null |
Default
| null | null |
null |
{
"abstract": " Program slicing provides explanations that illustrate how program outputs\nwere produced from inputs. We build on an approach introduced in prior work by\nPerera et al., where dynamic slicing was defined for pure higher-order\nfunctional programs as a Galois connection between lattices of partial inputs\nand partial outputs. We extend this approach to imperative functional programs\nthat combine higher-order programming with references and exceptions. We\npresent proofs of correctness and optimality of our approach and a\nproof-of-concept implementation and experimental evaluation.\n",
"title": "Imperative Functional Programs that Explain their Work"
}
| null | null |
[
"Computer Science"
] | null | true | null |
18020
| null |
Validated
| null | null |
null |
{
"abstract": " Synthetic dimensions alter one of the most fundamental properties in nature,\nthe dimension of space. They allow, for example, a real three-dimensional\nsystem to act as effectively four-dimensional. Driven by such possibilities,\nsynthetic dimensions have been engineered in ongoing experiments with ultracold\nmatter. We show that rotational states of ultracold molecules can be used as\nsynthetic dimensions extending to many - potentially hundreds of - synthetic\nlattice sites. Microwaves coupling rotational states drive fully controllable\nsynthetic inter-site tunnelings, enabling, for example, topological band\nstructures. Interactions leads to even richer behavior: when molecules are\nfrozen in a real space lattice with uniform synthetic tunnelings, dipole\ninteractions cause the molecules to aggregate to a narrow strip in the\nsynthetic direction beyond a critical interaction strength, resulting in a\nquantum string or a membrane, with an emergent condensate that lives on this\nstring or membrane. All these phases can be detected using measurements of\nrotational state populations.\n",
"title": "Synthetic dimensions in ultracold molecules: quantum strings and membranes"
}
| null | null |
[
"Physics"
] | null | true | null |
18021
| null |
Validated
| null | null |
null |
{
"abstract": " The managed-metabolism hypothesis suggests that a cooperation barrier must be\novercome if self-producing chemical organizations are to transition from\nnon-life to life. This barrier prevents un-managed, self-organizing,\nautocatalytic networks of molecular species from individuating into complex,\ncooperative organizations. The barrier arises because molecular species that\ncould otherwise make significant cooperative contributions to the success of an\norganization will often not be supported within the organization, and because\nside reactions and other free-riding processes will undermine cooperation. As a\nresult, the barrier seriously limits the possibility space that can be explored\nby un-managed organizations, impeding individuation, complex functionality and\nthe transition to life. The barrier can be overcome comprehensively by\nappropriate management which implements a system of evolvable constraints. The\nconstraints support beneficial co-operators and suppress free riders. In this\nway management can manipulate the chemical processes of an autocatalytic\norganization, producing novel processes that serve the interests of the\norganization as a whole and that could not arise and persist spontaneously in\nan un-managed chemical organization. Management self-organizes because it is\nable to capture some of the benefits that are produced when its management of\nan autocatalytic organization promotes beneficial cooperation. Selection\ntherefore favours the emergence of managers that take over and manage chemical\norganizations so as to overcome the cooperation barrier. The managed-metabolism\nhypothesis shows that if management is to overcome the cooperation barrier\ncomprehensively, its interventions must be digitally coded. In this way, the\nhypothesis accounts for the two-tiered structure of all living cells in which a\ndigitally-coded genetic apparatus manages an analogically-informed metabolism.\n",
"title": "Self-Organization and The Origins of Life: The Managed-Metabolism Hypothesis"
}
| null | null | null | null | true | null |
18022
| null |
Default
| null | null |
null |
{
"abstract": " Silicon drift detectors (SDDs) revolutionized spectroscopy in fields as\ndiverse as geology and dentistry. For a subset of experiments at ultra-fast,\nx-ray free-electron lasers (FELs), SDDs can make substantial contributions.\nOften the unknown spectrum is interesting, carrying science data, or the\nbackground measurement is useful to identify unexpected signals. Many\nmeasurements involve only several discrete photon energies known a priori. We\ndesigned a pulse function (a combination of gradual step and exponential decay\nfunction) and demonstrated that for individual pulses the signal amplitude,\npeaking time, and pulse amplitude are interrelated and the signal amplitude and\npeaking time are obtained for each pulse by fitting. Avoiding pulse shaping\nreduced peaking times to tens of nanoseconds, resulting in reduced pulse\npile-up and allowing decomposition of remaining pulse pile-up at photon\nseparation times down to 100~ns while yielding time-of-arrival information with\nprecision of 10~nanoseconds. At pulsed sources or high photon rates, photon\npile-up still occurs. We showed that the area of one photon peaks is not\nsuitable for estimating high photon rates while pile-up spectrum fitting is\nrelatively simple and preferable to pile-up spectrum deconvolution. We\ndeveloped a photon pile-up model for constant intensity sources, extended it to\nvariable intensity sources (typical for FELs) and used it to fit a complex\npile-up spectrum, demonstrating its accuracy. Based on the pile-up model, we\ndeveloped a Bayesian pile-up decomposition method that allows decomposing\npile-up of single events with up to 6 photons from 6 monochromatic lines with\n99% accuracy. The usefulness of SDDs will continue into the x-ray FEL era of\nscience. Their successors, the ePixS hybrid pixel detectors, already offer\nhundreds of pixels, each with similar performance to an SDD, in a compact,\nrobust and affordable package.\n",
"title": "Pile-up Reduction, Bayesian Decomposition and Applications of Silicon Drift Detectors at LCLS"
}
| null | null | null | null | true | null |
18023
| null |
Default
| null | null |
null |
{
"abstract": " Online learning to rank is a core problem in information retrieval and\nmachine learning. Many provably efficient algorithms have been recently\nproposed for this problem in specific click models. The click model is a model\nof how the user interacts with a list of documents. Though these results are\nsignificant, their impact on practice is limited, because all proposed\nalgorithms are designed for specific click models and lack convergence\nguarantees in other models. In this work, we propose BatchRank, the first\nonline learning to rank algorithm for a broad class of click models. The class\nencompasses two most fundamental click models, the cascade and position-based\nmodels. We derive a gap-dependent upper bound on the $T$-step regret of\nBatchRank and evaluate it on a range of web search queries. We observe that\nBatchRank outperforms ranked bandits and is more robust than CascadeKL-UCB, an\nexisting algorithm for the cascade model.\n",
"title": "Online Learning to Rank in Stochastic Click Models"
}
| null | null | null | null | true | null |
18024
| null |
Default
| null | null |
null |
{
"abstract": " We extend a technique called Compiling Control. The technique transforms\ncoroutining logic programs into logic programs that, when executed under the\nstandard left-to-right selection rule (and not using any delay features) have\nthe same computational behavior as the coroutining program. In recent work, we\nrevised Compiling Control and reformulated it as an instance of Abstract\nConjunctive Partial Deduction. This work was mostly focused on the program\nanalysis performed in Compiling Control. In the current paper, we focus on the\nsynthesis of the transformed program. Instead of synthesizing a new logic\nprogram, we synthesize a CHR(Prolog) program which mimics the coroutining\nprogram. The synthesis to CHR yields programs containing only simplification\nrules, which are particularly amenable to certain static analysis techniques.\nThe programs are also more concise and readable and can be ported to CHR\nimplementations embedded in other languages than Prolog.\n",
"title": "Transforming Coroutining Logic Programs into Equivalent CHR Programs"
}
| null | null | null | null | true | null |
18025
| null |
Default
| null | null |
null |
{
"abstract": " Large-scale vortices in protoplanetary disks are thought to form and survive\nfor long periods of time. Hence, they can significantly change the global disk\nevolution and particularly the distribution of the solid particles embedded in\nthe gas, possibly explaining asymmetries and dust concentrations recently\nobserved at sub-millimeter and millimeter wavelengths. We investigate the\nspatial distribution of dust grains using a simple model of protoplanetary disk\nhosted by a giant gaseous vortex. We explore the dependence of the results on\ngrain size and deduce possible consequences and predictions for observations of\nthe dust thermal emission at sub-millimeter and millimeter wavelengths. Global\n2D simulations with a bi-fluid code are used to follow the evolution of a\nsingle population of solid particles aerodynamically coupled to the gas.\nPossible observational signatures of the dust thermal emission are obtained\nusing simulators of ALMA and ngVLA observations. We find that a giant vortex\nnot only captures dust grains with Stokes number St < 1 but can also affect the\ndistribution of larger grains (with St '~' 1) carving a gap associated to a\nring composed of incompletely trapped particles. The results are presented for\ndifferent particle size and associated to their possible signatures in disk\nobservations. Gap clearing in the dust spatial distribution could be due to the\ninteraction with a giant gaseous vortex and their associated spiral waves,\nwithout the gravitational assistance of a planet. Hence, strong dust\nconcentrations at short sub-mm wavelengths associated with a gap and an\nirregular ring at longer mm and cm wavelengths could indicate the presence of\nan unseen gaseous vortex.\n",
"title": "Gap and rings carved by vortices in protoplanetary dust"
}
| null | null | null | null | true | null |
18026
| null |
Default
| null | null |
null |
{
"abstract": " This paper introduces a new nonlinear dictionary learning method for\nhistograms in the probability simplex. The method leverages optimal transport\ntheory, in the sense that our aim is to reconstruct histograms using so-called\ndisplacement interpolations (a.k.a. Wasserstein barycenters) between dictionary\natoms; such atoms are themselves synthetic histograms in the probability\nsimplex. Our method simultaneously estimates such atoms, and, for each\ndatapoint, the vector of weights that can optimally reconstruct it as an\noptimal transport barycenter of such atoms. Our method is computationally\ntractable thanks to the addition of an entropic regularization to the usual\noptimal transportation problem, leading to an approximation scheme that is\nefficient, parallel and simple to differentiate. Both atoms and weights are\nlearned using a gradient-based descent method. Gradients are obtained by\nautomatic differentiation of the generalized Sinkhorn iterations that yield\nbarycenters with entropic smoothing. Because of its formulation relying on\nWasserstein barycenters instead of the usual matrix product between dictionary\nand codes, our method allows for nonlinear relationships between atoms and the\nreconstruction of input data. We illustrate its application in several\ndifferent image processing settings.\n",
"title": "Wasserstein Dictionary Learning: Optimal Transport-based unsupervised non-linear dictionary learning"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
18027
| null |
Validated
| null | null |
null |
{
"abstract": " Smillie (1984) proved an interesting result on the stability of nonlinear,\ntime-invariant, strongly cooperative, and tridiagonal dynamical systems. This\nresult has found many applications in models from various fields including\nbiology, ecology, and chemistry. Smith (1991) has extended Smillie's result and\nproved entrainment in the case where the vector field is time-varying and\nperiodic. We use the theory of linear totally nonnegative differential systems\ndeveloped by Schwarz (1970) to give a generalization of these two results. This\nis based on weakening the requirement for strong cooperativity to\ncooperativity, and adding an additional observability-type condition.\n",
"title": "A Generalization of Smillie's Theorem on Strongly Cooperative Tridiagonal Systems"
}
| null | null | null | null | true | null |
18028
| null |
Default
| null | null |
null |
{
"abstract": " Geospatial semantics is a broad field that involves a variety of research\nareas. The term semantics refers to the meaning of things, and is in contrast\nwith the term syntactics. Accordingly, studies on geospatial semantics usually\nfocus on understanding the meaning of geographic entities as well as their\ncounterparts in the cognitive and digital world, such as cognitive geographic\nconcepts and digital gazetteers. Geospatial semantics can also facilitate the\ndesign of geographic information systems (GIS) by enhancing the\ninteroperability of distributed systems and developing more intelligent\ninterfaces for user interactions. During the past years, a lot of research has\nbeen conducted, approaching geospatial semantics from different perspectives,\nusing a variety of methods, and targeting different problems. Meanwhile, the\narrival of big geo data, especially the large amount of unstructured text data\non the Web, and the fast development of natural language processing methods\nenable new research directions in geospatial semantics. This chapter,\ntherefore, provides a systematic review on the existing geospatial semantic\nresearch. Six major research areas are identified and discussed, including\nsemantic interoperability, digital gazetteers, geographic information\nretrieval, geospatial Semantic Web, place semantics, and cognitive geographic\nconcepts.\n",
"title": "Geospatial Semantics"
}
| null | null | null | null | true | null |
18029
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we continue the study in \\cite{MiaoTX:DNLS:Stab}. We use the\nperturbation argument, modulational analysis and the energy argument in\n\\cite{MartelMT:Stab:gKdV, MartelMT:Stab:NLS} to show the stability of the sum\nof two solitary waves with weak interactions for the generalized derivative\nSchrödinger equation (gDNLS) in the energy space. Here (gDNLS) hasn't the\nGalilean transformation invariance, the pseudo-conformal invariance and the\ngauge transformation invariance, and the case $\\sigma>1$ we considered\ncorresponds to the $L^2$-supercritical case.\n",
"title": "Stability of the sum of two solitary waves for (gDNLS) in the energy space"
}
| null | null | null | null | true | null |
18030
| null |
Default
| null | null |
null |
{
"abstract": " This technical note describes a new baseline for the Natural Questions. Our\nmodel is based on BERT and reduces the gap between the model F1 scores reported\nin the original dataset paper and the human upper bound by 30% and 50% relative\nfor the long and short answer tasks respectively. This baseline has been\nsubmitted to the official NQ leaderboard at\nai.google.com/research/NaturalQuestions and we plan to opensource the code for\nit in the near future.\n",
"title": "A BERT Baseline for the Natural Questions"
}
| null | null | null | null | true | null |
18031
| null |
Default
| null | null |
null |
{
"abstract": " Even as we advance the frontiers of physics knowledge, our understanding of\nhow this knowledge evolves remains at the descriptive levels of Popper and\nKuhn. Using the APS publications data sets, we ask in this letter how new\nknowledge is built upon old knowledge. We do so by constructing year-to-year\nbibliographic coupling networks, and identify in them validated communities\nthat represent different research fields. We then visualize their evolutionary\nrelationships in the form of alluvial diagrams, and show how they remain intact\nthrough APS journal splits. Quantitatively, we see that most fields undergo\nweak Popperian mixing, and it is rare for a field to remain isolated/undergo\nstrong mixing. The sizes of fields obey a simple linear growth with\nrecombination. We can also reliably predict the merging between two fields, but\nnot for the considerably more complex splitting. Finally, we report a case\nstudy of two fields that underwent repeated merging and splitting around 1995,\nand how these Kuhnian events are correlated with breakthroughs on BEC, quantum\nteleportation, and slow light. This impact showed up quantitatively in the\ncitations of the BEC field as a larger proportion of references from during and\nshortly after these events.\n",
"title": "Knowledge Evolution in Physics Research: An Analysis of Bibliographic Coupling Networks"
}
| null | null | null | null | true | null |
18032
| null |
Default
| null | null |
null |
{
"abstract": " We study the energy functional on the set of Lagrangian tori in\n$\\mathbb{C}P^2$ . We prove that the value of the energy functional on a certain\nfamily of Hamiltonian minimal Lagrangian tori in $\\mathbb{C}P^2$ is strictly\nlarger than energy of the Clifford torus.\n",
"title": "On a lower bound for the energy functional on a family of Hamiltonian minimal Lagrangian tori in $\\mathbb{C}P^2$"
}
| null | null |
[
"Mathematics"
] | null | true | null |
18033
| null |
Validated
| null | null |
null |
{
"abstract": " Viewing the trajectory of a patient as a dynamical system, a recurrent neural\nnetwork was developed to learn the course of patient encounters in the\nPediatric Intensive Care Unit (PICU) of a major tertiary care center. Data\nextracted from Electronic Medical Records (EMR) of about 12000 patients who\nwere admitted to the PICU over a period of more than 10 years were leveraged.\nThe RNN model ingests a sequence of measurements which include physiologic\nobservations, laboratory results, administered drugs and interventions, and\ngenerates temporally dynamic predictions for in-ICU mortality at user-specified\ntimes. The RNN's ICU mortality predictions offer significant improvements over\nthose from two clinically-used scores and static machine learning algorithms.\n",
"title": "Dynamic Mortality Risk Predictions in Pediatric Critical Care Using Recurrent Neural Networks"
}
| null | null | null | null | true | null |
18034
| null |
Default
| null | null |
null |
{
"abstract": " We consider the Schrödinger equation on a half space in any dimension with\na class of nonhomogeneous boundary conditions including Dirichlet, Neuman and\nthe so-called transparent boundary conditions. Building upon recent local in\ntime Strichartz estimates (for Dirichlet boundary conditions), we obtain global\nStrichartz estimates for initial data in $H^s,\\ 0\\leq s\\leq 2$ and boundary\ndata in a natural space $\\mathcal{H}^s$. For $s\\geq 1/2$, the issue of\ncompatibility conditions requires a thorough analysis of the $\\mathcal{H}^s$\nspace. As an application we solve nonlinear Schrödinger equations and\nconstruct global asymptotically linear solutions for small data. A discussion\nis included on the appropriate notion of scattering in this framework, and the\noptimality of the $\\mathcal{H}^s$ space.\n",
"title": "Global Strichartz estimates for the Schrödinger equation with non zero boundary conditions and applications"
}
| null | null | null | null | true | null |
18035
| null |
Default
| null | null |
null |
{
"abstract": " Background\nThe nuclear structure of the cluster bands in $^{20}$Ne presents a challenge\nfor different theoretical approaches. It is especially difficult to explain the\nbroad 0$^+$, 2$^+$ states at 9 MeV excitation energy. Simultaneously, it is\nimportant to obtain more reliable experimental data for these levels in order\nto quantitatively assess the theoretical framework.\nPurpose\nTo obtain new data on $^{20}$Ne $\\alpha$ cluster structure. Method Thick\ntarget inverse kinematics technique was used to study the $^{16}$O+$\\alpha$\nresonance elastic scattering and the data were analyzed using an \\textit{R}\nmatrix approach. The $^{20}$Ne spectrum, the cluster and nucleon spectroscopic\nfactors were calculated using cluster-nucleon configuration interaction model\n(CNCIM).\nResults\nWe determined the parameters of the broad resonances in\n\\textsuperscript{20}Ne: 0$^+$ level at 8.77 $\\pm$ 0.150 MeV with a width of 750\n(+500/-220) keV; 2$^+$ level at 8.75 $\\pm$ 0.100 MeV with the width of 695\n$\\pm$ 120 keV; the width of 9.48 MeV level of 65 $\\pm$ 20 keV and showed that\n9.19 MeV, 2$^+$ level (if exists) should have width $\\leq$ 10 keV. The detailed\ncomparison of the theoretical CNCIM predictions with the experimental data on\ncluster states was made.\nConclusions\nOur experimental results by the TTIK method generally confirm the adopted\ndata on $\\alpha$ cluster levels in $^{20}$Ne. The CNCIM gives a good\ndescription of the $^{20}$Ne positive parity states up to an excitation energy\nof $\\sim$ 7 MeV, predicting reasonably well the excitation energy of the states\nand their cluster and single particle properties. At higher excitations, the\nqualitative disagreement with the experimentally observed structure is evident,\nespecially for broad resonances.\n",
"title": "Structure of $^{20}$Ne states in the resonance $^{16}$O+$α$ elastic scattering"
}
| null | null | null | null | true | null |
18036
| null |
Default
| null | null |
null |
{
"abstract": " We investigate the terminal-pairibility problem in the case when the base\ngraph is a complete bipartite graph, and the demand graph is also bipartite\nwith the same color classes. We improve the lower bound on maximum value of\n$\\Delta(D)$ which still guarantees that the demand graph $D$ is\nterminal-pairable in this setting. We also prove a sharp theorem on the maximum\nnumber of edges such a demand graph can have.\n",
"title": "Terminal-Pairability in Complete Bipartite Graphs"
}
| null | null | null | null | true | null |
18037
| null |
Default
| null | null |
null |
{
"abstract": " Meta learning of optimal classifier error rates allows an experimenter to\nempirically estimate the intrinsic ability of any estimator to discriminate\nbetween two populations, circumventing the difficult problem of estimating the\noptimal Bayes classifier. To this end we propose a weighted nearest neighbor\n(WNN) graph estimator for a tight bound on the Bayes classification error; the\nHenze-Penrose (HP) divergence. Similar to recently proposed HP estimators\n[berisha2016], the proposed estimator is non-parametric and does not require\ndensity estimation. However, unlike previous approaches the proposed estimator\nis rate-optimal, i.e., its mean squared estimation error (MSEE) decays to zero\nat the fastest possible rate of $O(1/M+1/N)$ where $M,N$ are the sample sizes\nof the respective populations. We illustrate the proposed WNN meta estimator\nfor several simulated and real data sets.\n",
"title": "Rate-optimal Meta Learning of Classification Error"
}
| null | null | null | null | true | null |
18038
| null |
Default
| null | null |
null |
{
"abstract": " Event detection is a critical feature in data-driven systems as it assists\nwith the identification of nominal and anomalous behavior. Event detection is\nincreasingly relevant in robotics as robots operate with greater autonomy in\nincreasingly unstructured environments. In this work, we present an accurate,\nrobust, fast, and versatile measure for skill and anomaly identification. A\ntheoretical proof establishes the link between the derivative of the\nlog-likelihood of the HMM filtered belief state and the latest emission\nprobabilities. The key insight is the inverse relationship in which gradient\nanalysis is used for skill and anomaly identification. Our measure showed\nbetter performance across all metrics than related state-of-the art works. The\nresult is broadly applicable to domains that use HMMs for event detection.\n",
"title": "Fast, Robust, and Versatile Event Detection through HMM Belief State Gradient Measures"
}
| null | null | null | null | true | null |
18039
| null |
Default
| null | null |
null |
{
"abstract": " This paper locally classifies finite-dimensional Lie algebras of conformal\nand Killing vector fields on $\\mathbb{R}^2$ relative to an arbitrary\npseudo-Riemannian metric. Several results about their geometric properties are\ndetailed, e.g. their invariant distributions and induced symplectic structures.\nFindings are illustrated with two examples of physical nature: the\nMilne--Pinney equation and the projective Schrödinger equation on the Riemann\nsphere.\n",
"title": "Geometric features of Vessiot--Guldberg Lie algebras of conformal and Killing vector fields on $\\mathbb{R}^2$"
}
| null | null | null | null | true | null |
18040
| null |
Default
| null | null |
null |
{
"abstract": " This work sets out to compute and discuss effects of spin, velocity and\ndimensionality on inter-particle potentials systematically derived from gauge\nfield-theoretic models. We investigate the interaction of fermionic particles\nby the exchange of a vector field in a parity-preserving description in\nfive-dimensional $(5D)$ space-time. A particular dimensional reduction\nprescription is adopted $-$ reduction by dimensional restriction $-$ and\nspecial effects, like a pseudo-spin dependence, show up in four dimensions\n$(4D)$. What we refer to as pseudo-spin shall be duly explained. The main idea\nwe try to convey is that the calculation of the potentials in $5D$ and the\nconsequent reduction to $4D$ exhibits new effects that are not present if the\npotential is calculated in $4D$ after the action has been reduced.\n",
"title": "Fermion inter-particle potentials in 5D and a dimensional restriction prescription to 4D"
}
| null | null | null | null | true | null |
18041
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we provide an axiomatic characterization of the idempotent\ndiscrete uninorms by means of three conditions only: conservativeness,\nsymmetry, and nondecreasing monotonicity. We also provide an alternative\ncharacterization involving the bisymmetry property. Finally, we provide a\ngraphical characterization of these operations in terms of their contour plots,\nand we mention a few open questions for further research.\n",
"title": "Characterizations of idempotent discrete uninorms"
}
| null | null | null | null | true | null |
18042
| null |
Default
| null | null |
null |
{
"abstract": " We investigate the constraints on the sum of neutrino masses ($\\Sigma m_\\nu$)\nusing the most recent cosmological data, which combines the distance\nmeasurement from baryonic acoustic oscillation in the extended Baryon\nOscillation Spectroscopic Survey DR14 quasar sample with the power spectra of\ntemperature and polarization anisotropies in the cosmic microwave background\nfrom the Planck 2015 data release. We also use other low-redshift observations\nincluding the baryonic acoustic oscillation at relatively low redshifts, the\nsupernovae of type Ia and the local measurement of Hubble constant. In the\nstandard cosmological constant $\\Lambda$ cold dark matter plus massive neutrino\nmodel, we obtain the $95\\%$ \\acl{CL} upper limit to be $\\Sigma\nm_\\nu<0.129~\\mathrm{eV}$ for the degenerate mass hierarchy, $\\Sigma\nm_{\\nu}<0.159~\\mathrm{eV}$ for the normal mass hierarchy, and $\\Sigma\nm_{\\nu}<0.189~\\mathrm{eV}$ for the inverted mass hierarchy. Based on Bayesian\nevidence, we find that the degenerate hierarchy is positively supported, and\nthe current data combination can not distinguish normal and inverted\nhierarchies. Assuming the degenerate mass hierarchy, we extend our study to\nnon-standard cosmological models including the generic dark energy, the spatial\ncurvature, and the extra relativistic degrees of freedom, respectively, but\nfind these models not favored by the data.\n",
"title": "Constraints on the sum of neutrino masses using cosmological data including the latest extended Baryon Oscillation Spectroscopic Survey DR14 quasar sample"
}
| null | null | null | null | true | null |
18043
| null |
Default
| null | null |
null |
{
"abstract": " We show that the number of unique function mappings in a neural network\nhypothesis space is inversely proportional to $\\prod_lU_l!$, where $U_{l}$ is\nthe number of neurons in the hidden layer $l$.\n",
"title": "The effect of the choice of neural network depth and breadth on the size of its hypothesis space"
}
| null | null | null | null | true | null |
18044
| null |
Default
| null | null |
null |
{
"abstract": " Black-box explanation is the problem of explaining how a machine learning\nmodel -- whose internal logic is hidden to the auditor and generally complex --\nproduces its outcomes. Current approaches for solving this problem include\nmodel explanation, outcome explanation as well as model inspection. While these\ntechniques can be beneficial by providing interpretability, they can be used in\na negative manner to perform fairwashing, which we define as promoting the\nperception that a machine learning model respects some ethical values while it\nmight not be the case. In particular, we demonstrate that it is possible to\nsystematically rationalize decisions taken by an unfair black-box model using\nthe model explanation as well as the outcome explanation approaches with a\ngiven fairness metric. Our solution, LaundryML, is based on a regularized rule\nlist enumeration algorithm whose objective is to search for fair rule lists\napproximating an unfair black-box model. We empirically evaluate our\nrationalization technique on black-box models trained on real-world datasets\nand show that one can obtain rule lists with high fidelity to the black-box\nmodel while being considerably less unfair at the same time.\n",
"title": "Fairwashing: the risk of rationalization"
}
| null | null | null | null | true | null |
18045
| null |
Default
| null | null |
null |
{
"abstract": " We derive the finite-volume correction to the binding energy of an N-particle\nquantum bound state in a cubic periodic volume. Our results are applicable to\nbound states with arbitrary composition and total angular momentum, and in any\nnumber of spatial dimensions. The only assumptions are that the interactions\nhave finite range. The finite-volume correction is a sum of contributions from\nall possible breakup channels. In the case where the separation is into two\nbound clusters, our result gives the leading volume dependence up to\nexponentially small corrections. If the separation is into three or more\nclusters, there is a power-law factor that is beyond the scope of this work,\nhowever our result again determines the leading exponential dependence. We also\npresent two independent methods that use finite-volume data to determine\nasymptotic normalization coefficients. The coefficients are useful to determine\nlow-energy capture reactions into weakly bound states relevant for nuclear\nastrophysics. Using the techniques introduced here, one can even extract the\ninfinite-volume energy limit using data from a single-volume calculation. The\nderived relations are tested using several exactly solvable systems and\nnumerical examples. We anticipate immediate applications to lattice\ncalculations of hadronic, nuclear, and cold atomic systems.\n",
"title": "Volume Dependence of N-Body Bound States"
}
| null | null | null | null | true | null |
18046
| null |
Default
| null | null |
null |
{
"abstract": " Additive models, such as produced by gradient boosting, and full interaction\nmodels, such as classification and regression trees (CART), are widely used\nalgorithms that have been investigated largely in isolation. We show that these\nmodels exist along a spectrum, revealing never-before-known connections between\nthese two approaches. This paper introduces a novel technique called\ntree-structured boosting for creating a single decision tree, and shows that\nthis method can produce models equivalent to CART or gradient boosted stumps at\nthe extremes by varying a single parameter. Although tree-structured boosting\nis designed primarily to provide both the model interpretability and predictive\nperformance needed for high-stake applications like medicine, it also can\nproduce decision trees represented by hybrid models between CART and boosted\nstumps that can outperform either of these approaches.\n",
"title": "Tree-Structured Boosting: Connections Between Gradient Boosted Stumps and Full Decision Trees"
}
| null | null | null | null | true | null |
18047
| null |
Default
| null | null |
null |
{
"abstract": " We formulate a quasiclassical theory ($\\omega_c\\tau \\lesssim 1$ with\n$\\omega_c$ as the cyclotron frequency and $\\tau$ as the relaxation time) to\nstudy the influence of magnetic field on electron-impurity scattering process\nin the two-dimensional electron gas. We introduce a general recipe based on an\nabstraction of the detailed impurity scattering process to define the\nscattering parameter such as the incoming and outgoing momentum and coordinate\njump. In this picture, we can conveniently describe the skew scattering and\ncoordinate jump, which will eventually modify the Boltzmann equation. We find\nan anomalous Hall resistivity different from the conventional Boltzmann-Drude\nresult and a negative magnetoresistivity parabolic in magnetic field. The\norigin of these results has been analyzed. The relevance between our theory and\nrecent simulation and experimental works is also discussed. Our theory\ndominates in dilute impurity system where the correlation effect is negligible.\n",
"title": "Magnetic field influenced electron-impurity scattering and magnetotransport"
}
| null | null | null | null | true | null |
18048
| null |
Default
| null | null |
null |
{
"abstract": " Context. The Rosetta Orbiter Spectrometer for Ion and Neutral Analysis\n(ROSINA) was designed to measure the composition of the gas in the coma of\ncomet 67P/Churyumov-Gerasimenko, the target of the European Space Agency's\nRosetta mission. In addition to the volatiles, ROSINA measured refractories\nsputtered off the comet by the interaction of solar wind protons with the\nsurface of the comet.\nAims. The origin of different solar system materials is still heavily\ndebated. Isotopic ratios can be used to distinguish between different\nreservoirs and investigate processes occurring during the formation of the\nsolar system.\nMethods. ROSINA consisted of two mass spectrometers and a pressure sensor. In\nthe ROSINA Double Focusing Mass Spectrometer (DFMS), the neutral gas of\ncometary origin was ionized and then deflected in an electric and a magnetic\nfield that separated the ions based on their mass-to-charge ratio. The DFMS had\na high mass resolution, dynamic range, and sensitivity that allowed detection\nof rare species and the known major volatiles.\nResults. We measured the relative abundance of all three stable silicon\nisotopes with the ROSINA instrument on board the Rosetta spacecraft.\nFurthermore, we measured $^{13}$C/$^{12}$C in C$_2$H$_4$, C$_2$H$_5$, and CO.\nThe DFMS in situ measurements indicate that the average silicon isotopic\ncomposition shows depletion in the heavy isotopes $^{29}$Si and $^{30}$Si with\nrespect to $^{28}$Si and solar abundances, while $^{13}$C to $^{12}$C is\nanalytically indistinguishable from bulk planetary and meteorite compositions.\nAlthough the origin of the deficiency of the heavy silicon isotopes cannot be\nexplained unambiguously, we discuss mechanisms that could have contributed to\nthe measured depletion of the isotopes $^{29}$Si and $^{30}$Si.\n",
"title": "Evidence for depletion of heavy silicon isotopes at comet 67P/Churyumov-Gerasimenko"
}
| null | null | null | null | true | null |
18049
| null |
Default
| null | null |
null |
{
"abstract": " Walking quadruped robots face challenges in positioning their feet and\nlifting their legs during gait cycles over uneven terrain. The robot Laika is\nunder development as a quadruped with a flexible, actuated spine designed to\nassist with foot movement and balance during these gaits. This paper presents\nthe first set of hardware designs for the spine of Laika, a physical prototype\nof those designs, and tests in both hardware and simulations that show the\nprototype's capabilities. Laika's spine is a tensegrity structure, used for its\nadvantages with weight and force distribution, and represents the first working\nprototype of a tensegrity spine for a quadruped robot. The spine bends by\nadjusting the lengths of the cables that separate its vertebrae, and twists\nusing an actuated rotating vertebra at its center. The current prototype of\nLaika has stiff legs attached to the spine, and is used as a test setup for\nevaluation of the spine itself. This work shows the advantages of Laika's spine\nby demonstrating the spine lifting each of the robot's four feet, both as a\nform of balancing and as a precursor for a walking gait. These foot motions,\nusing specific combinations of bending and rotation movements of the spine, are\nmeasured in both simulation and hardware experiments. Hardware data are used to\ncalibrate the simulations, such that the simulations can be used for control of\nbalancing or gait cycles in the future. Future work will attach actuated legs\nto Laika's spine, and examine balancing and gait cycles when combined with leg\nmovements.\n",
"title": "Design, Simulation, and Testing of a Flexible Actuated Spine for Quadruped Robots"
}
| null | null | null | null | true | null |
18050
| null |
Default
| null | null |
null |
{
"abstract": " As the field of neuroimaging grows, it can be difficult for scientists within\nthe field to gain and maintain a detailed understanding of its ever-changing\nlandscape. While collaboration and citation networks highlight important\ncontributions within the field, the roles of and relations among specific areas\nof study can remain quite opaque. Here, we apply techniques from network\nscience to map the landscape of neuroimaging research documented in the journal\nNeuroImage over the past decade. We create a network in which nodes represent\nresearch topics, and edges give the degree to which these topics tend to be\ncovered in tandem. The network displays small-world architecture, with\ncommunities characterized by common imaging modalities and medical\napplications, and with bridges that integrate these distinct subfields. Using\nnode-level analysis, we quantify the structural roles of individual topics\nwithin the neuroimaging landscape, and find high levels of clustering within\nthe structural MRI subfield as well as increasing participation among topics\nrelated to psychiatry. The overall prevalence of a topic is unrelated to the\nprevalence of its neighbors, but the degree to which a topic becomes more or\nless popular over time is strongly related to changes in the prevalence of its\nneighbors. Broadly, this work presents a cohesive model for understanding the\nlandscape of neuroimaging research across the field, in broad subfields, and\nwithin specific topic areas.\n",
"title": "The landscape of NeuroImage-ing research"
}
| null | null | null | null | true | null |
18051
| null |
Default
| null | null |
null |
{
"abstract": " The number of trees T in the random forest (RF) algorithm for supervised\nlearning has to be set by the user. It is controversial whether T should simply\nbe set to the largest computationally manageable value or whether a smaller T\nmay in some cases be better. While the principle underlying bagging is that\n\"more trees are better\", in practice the classification error rate sometimes\nreaches a minimum before increasing again for increasing number of trees. The\ngoal of this paper is four-fold: (i) providing theoretical results showing that\nthe expected error rate may be a non-monotonous function of the number of trees\nand explaining under which circumstances this happens; (ii) providing\ntheoretical results showing that such non-monotonous patterns cannot be\nobserved for other performance measures such as the Brier score and the\nlogarithmic loss (for classification) and the mean squared error (for\nregression); (iii) illustrating the extent of the problem through an\napplication to a large number (n = 306) of datasets from the public database\nOpenML; (iv) finally arguing in favor of setting it to a computationally\nfeasible large number, depending on convergence properties of the desired\nperformance measure.\n",
"title": "To tune or not to tune the number of trees in random forest?"
}
| null | null | null | null | true | null |
18052
| null |
Default
| null | null |
null |
{
"abstract": " This study investigates short-crested wave breaking over a planar beach by\nusing the mesh-free Smoothed Particle Hydrodynamics model, GPUSPH. The\nshort-crested waves are created by generating intersecting wave trains in a\nnumerical wave basin. We examine the influence of beach slope, incident wave\nheight, and incident wave angle on the generated short-crested waves.\nShort-crested wave breaking over a steeper beach generates stronger rip\ncurrents, and larger circulation cells in front of the beach. Intersecting wave\ntrains with a larger incident wave height drive a more complicated\nshort-crested wave field including isolated breakers and wave amplitude\ndiffraction. Nearshore circulation induced by short-crested wave breaking is\ngreatly influenced by the incident wave angle (or the rip current spacing).\nThere is no secondary circulation cell between the nodal line and the antinodal\nline if the rip current spacing is narrow. However, there are multiple\nsecondary circulation cells observed when the rip current spacing is relatively\nlarge.\n",
"title": "SPH Modeling of Short-crested Waves"
}
| null | null | null | null | true | null |
18053
| null |
Default
| null | null |
null |
{
"abstract": " Transition metal dichalcogenides (TMDCs), with their two-dimensional\nstructures and sizable bandgaps, are good candidates for barrier materials in\ntunneling field-effect transistor (TFET) formed from atomic precision vertical\nstacks of graphene and insulating crystals of a few atomic layers in thickness.\nWe report first-principles study of the electronic properties of the\nGraphene/WS$_2$/Graphene sandwich structure revealing strong interface effects\non dielectric properties and predicting a high ON/OFF ratio with an appropriate\nWS$_2$ thickness and a suitable range of the gate voltage. Both the band\nspin-orbit coupling splitting and the dielectric constant of the WS$_2$ layer\ndepend on its thickness when in contact with the graphene electrodes,\nindicating strong influence from graphene across the interfaces. The dielectric\nconstant is significantly reduced from the bulk WS$_2$ value. The effective\nbarrier height varies with WS$_2$ thickness and can be tuned by a gate voltage.\nThese results are critical for future nanoelectronic device designs.\n",
"title": "Tunneling Field-Effect Junctions with WS$_2$ barrier"
}
| null | null |
[
"Physics"
] | null | true | null |
18054
| null |
Validated
| null | null |
null |
{
"abstract": " We theoretically study bilayer superconducting topological insulator film, in\nwhich superconductivity exists for both top and bottom surface states. We show\nthat an in-plane magnetic field can drive the system into Larkin-Ovchinnikov\n(LO) phase, where electrons are paired with finite momenta. The LO phase is\ntopologically non-trivial and characterized by a Z 2 topological invariant,\nleading to a Majorana zero mode chain along the edge perpendicular to in-plane\nmagnetic fields.\n",
"title": "Topological Larkin-Ovchinnikov phase and Majorana zero mode chain in bilayer superconducting topological insulator films"
}
| null | null | null | null | true | null |
18055
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we consider a three-node cooperative wireless powered\ncommunication system consisting of a multi-antenna hybrid access point (H-AP)\nand a single-antenna relay and a single-antenna user. The energy constrained\nrelay and user first harvest energy in the downlink and then the relay assists\nthe user using the harvested power for information transmission in the uplink.\nThe optimal energy beamforming vector and the time split between harvest and\ncooperation are investigated. To reduce the computational complexity,\nsuboptimal designs are also studied, where closed-form expressions are derived\nfor the energy beamforming vector and the time split. For comparison purposes,\nwe also present a detailed performance analysis in terms of the achievable\noutage probability and the average throughput of an intuitive energy\nbeamforming scheme, where the H-AP directs all the energy towards the user. The\nfindings of the paper suggest that implementing multiple antennas at the H-AP\ncan significantly improve the system performance, and the closed-form\nsuboptimal energy beamforming vector and time split yields near optimal\nperformance. Also, for the intuitive beamforming scheme, a diversity order of\n(N+1)/2 can be achieved, where N is the number of antennas at the H-AP.\n",
"title": "Optimization and Analysis of Wireless Powered Multi-antenna Cooperative Systems"
}
| null | null |
[
"Computer Science"
] | null | true | null |
18056
| null |
Validated
| null | null |
null |
{
"abstract": " Let $F$ be a non-archimedean locally compact field. We study a class of\nLanglands-Shahidi pairs $({\\bf H},{\\bf L})$, consisting of a quasi-split\nconnected reductive group $\\bf H$ over $F$ and a Levi subgroup $\\bf L$ which is\nclosely related to a product of restriction of scalars of ${\\rm GL}_1$'s or\n${\\rm GL}_2$'s. We prove the compatibility of the resulting local factors with\nthe Langlands correspondence. In particular, let $E$ be a cubic separable\nextension of $F$. We consider a simply connected quasi-split semisimple group\n$\\bf H$ over $F$ of type $D_4$, with triality corresponding to $E$, and let\n$\\bf L$ be its Levi subgroup with derived group ${\\rm Res}_{E/F} {\\rm SL}_2$.\nIn this way we obtain Asai cube local factors attached to irreducible smooth\nrepresentations of ${\\rm GL}_2(E)$; we prove that they are Weil-Deligne factors\nobtained via the local Langlands correspondence for ${\\rm GL}_2(E)$ and tensor\ninduction from $E$ to $F$. A consequence is that Asai cube $\\gamma$- and\n$\\varepsilon$-factors become stable under twists by highly ramified characters.\n",
"title": "Asai cube L-functions and the local Langlands conjecture"
}
| null | null | null | null | true | null |
18057
| null |
Default
| null | null |
null |
{
"abstract": " Game semantics is a rich and successful class of denotational models for\nprogramming languages. Most game models feature a rather intuitive setup, yet\nsurprisingly difficult proofs of such basic results as associativity of\ncomposition of strategies. We set out to unify these models into a basic\nabstract framework for game semantics, game settings. Our main contribution is\nthe generic construction, for any game setting, of a category of games and\nstrategies. Furthermore, we extend the framework to deal with innocence, and\nprove that innocent strategies form a subcategory. We finally show that our\nconstructions cover many concrete cases, mainly among the early models and the\nvery recent sheaf-based ones.\n",
"title": "What's in a game? A theory of game models"
}
| null | null | null | null | true | null |
18058
| null |
Default
| null | null |
null |
{
"abstract": " We present a proposal for applying nanoscale magnetometry to the search for\nmagnetic monopoles in the spin ice materials holmium and dysprosium titanate.\nEmploying Monte Carlo simulations of the dipolar spin ice model, we find that\nwhen cooled to below $1.5\\,$K these materials exhibit a sufficiently low\nmonopole density to enable the direct observation of magnetic fields from\nindividual monopoles. At these temperatures we demonstrate that noise\nspectroscopy can capture the intrinsic fluctuations associated with monopole\ndynamics, allowing one to isolate the qualitative effects associated with both\nthe Coulomb interaction between monopoles and the topological constraints\nimplied by Dirac strings. We describe in detail three different nanoscale\nmagnetometry platforms (muon spin rotation, nitrogen vacancy defects, and\nnanoSQUID arrays) that can be used to detect monopoles in these experiments,\nand analyze the advantages of each.\n",
"title": "Proposal for the Detection of Magnetic Monopoles in Spin Ice via Nanoscale Magnetometry"
}
| null | null | null | null | true | null |
18059
| null |
Default
| null | null |
null |
{
"abstract": " Preprocessing tools for automated text analysis have become more widely\navailable in major languages, but non-English tools are often still limited in\ntheir functionality. When working with Spanish-language text, researchers can\neasily find tools for tokenization and stemming, but may not have the means to\nextract more complex word features like verb tense or mood. Yet Spanish is a\nmorphologically rich language in which such features are often identifiable\nfrom word form. Conjugation rules are consistent, but many special verbs and\nnouns take on different rules. While building a complete dictionary of known\nwords and their morphological rules would be labor intensive, resources to do\nso already exist, in spell checkers designed to generate valid forms of known\nwords. This paper introduces a set of tools for Spanish-language morphological\nanalysis, built using the COES spell checking tools, to label person, mood,\ntense, gender and number, derive a word's root noun or verb infinitive, and\nconvert verbs to their nominal form.\n",
"title": "Rule-Based Spanish Morphological Analyzer Built From Spell Checking Lexicon"
}
| null | null | null | null | true | null |
18060
| null |
Default
| null | null |
null |
{
"abstract": " We propose a new method for embedding graphs while preserving directed edge\ninformation. Learning such continuous-space vector representations (or\nembeddings) of nodes in a graph is an important first step for using network\ninformation (from social networks, user-item graphs, knowledge bases, etc.) in\nmany machine learning tasks.\nUnlike previous work, we (1) explicitly model an edge as a function of node\nembeddings, and we (2) propose a novel objective, the \"graph likelihood\", which\ncontrasts information from sampled random walks with non-existent edges.\nIndividually, both of these contributions improve the learned representations,\nespecially when there are memory constraints on the total size of the\nembeddings. When combined, our contributions enable us to significantly improve\nthe state-of-the-art by learning more concise representations that better\npreserve the graph structure.\nWe evaluate our method on a variety of link-prediction task including social\nnetworks, collaboration networks, and protein interactions, showing that our\nproposed method learn representations with error reductions of up to 76% and\n55%, on directed and undirected graphs. In addition, we show that the\nrepresentations learned by our method are quite space efficient, producing\nembeddings which have higher structure-preserving accuracy but are 10 times\nsmaller.\n",
"title": "Learning Edge Representations via Low-Rank Asymmetric Projections"
}
| null | null | null | null | true | null |
18061
| null |
Default
| null | null |
null |
{
"abstract": " We develop a feedback control method for networked epidemic spreading\nprocesses. In contrast to most prior works which consider mean field, open-loop\ncontrol schemes, the present work develops a novel framework for feedback\ncontrol of epidemic processes which leverages incomplete observations of the\nstochastic epidemic process in order to control the exact dynamics of the\nepidemic outbreak. We develop an observation model for the epidemic process,\nand demonstrate that if the set of observed nodes is sufficiently well\nstructured, then the random variables which denote the process' infections are\nconditionally independent given the observations. We then leverage the attained\nconditional independence property to construct tractable mechanisms for the\ninference and prediction of the process state, avoiding the need to use mean\nfield approximations or combinatorial representations. We conclude by\nformulating a one-step lookahead controller for the discrete-time\nSusceptible-Infected-Susceptible (SIS) epidemic process which leverages the\ndeveloped Bayesian inference and prediction mechanisms, and causes the epidemic\nto die out at a chosen rate.\n",
"title": "Inference, Prediction, and Control of Networked Epidemics"
}
| null | null | null | null | true | null |
18062
| null |
Default
| null | null |
null |
{
"abstract": " We prove that the number of iterations taken by the Weisfeiler-Leman\nalgorithm for configurations coming from Schreier graphs is closely linked to\nthe diameter of the graphs themselves: an upper bound is found for general\nSchreier graphs, and a lower bound holds for particular cases, such as for\nSchreier graphs with $G=\\mbox{SL}_{n}({\\mathbb F}_{q})$ ($q>2$) acting on\n$k$-tuples of vectors in ${\\mathbb F}_{q}^{n}$; moreover, an exact expression\nis found in the case of Cayley graphs.\n",
"title": "The Weisfeiler-Leman algorithm and the diameter of Schreier graphs"
}
| null | null |
[
"Mathematics"
] | null | true | null |
18063
| null |
Validated
| null | null |
null |
{
"abstract": " We prove an analogue of the Hardy-Littlewood conjecture on the asymptotic\ndistribution of prime constellations in the setting of short intervals in\nfunction fields of smooth projective curves over finite fields.\n",
"title": "Correlations between primes in short intervals on curves over finite fields"
}
| null | null | null | null | true | null |
18064
| null |
Default
| null | null |
null |
{
"abstract": " Besides enabling an enhanced mobile broadband, next generation of mobile\nnetworks (5G) are envisioned for the support of massive connectivity of\nheterogeneous Internet of Things (IoT)s. These IoTs are envisioned for a large\nnumber of use-cases including smart cities, environment monitoring, smart\nvehicles, etc. Unfortunately, most IoTs have very limited computing and storage\ncapabilities and need cloud services. Hence, connecting these devices through\n5G systems requires huge spectrum resources in addition to handling the massive\nconnectivity and improved security. This article discusses the challenges\nfacing the support of IoTs through 5G systems. The focus is devoted to\ndiscussing physical layer limitations in terms of spectrum resources and radio\naccess channel connectivity. We show how sparsity can be exploited for\naddressing these challenges especially in terms of enabling wideband spectrum\nmanagement and handling the connectivity by exploiting device-to-device\ncommunications and edge-cloud. Moreover, we identify major open problems and\nresearch directions that need to be explored towards enabling the support of\nmassive heterogeneous IoTs through 5G systems.\n",
"title": "Extracting and Exploiting Inherent Sparsity for Efficient IoT Support in 5G: Challenges and Potential Solutions"
}
| null | null | null | null | true | null |
18065
| null |
Default
| null | null |
null |
{
"abstract": " In this work, we prove the existence of local convex solution to the\ndegenerate Hessian equation\n",
"title": "Existence and convexity of local solutions to degenerate hessian equations"
}
| null | null | null | null | true | null |
18066
| null |
Default
| null | null |
null |
{
"abstract": " We prove the following superexponential distribution inequality: for any\nintegrable $g$ on $[0,1)^{d}$ with zero average, and any $\\lambda>0$ \\[ |\\{ x\n\\in [0,1)^{d} \\; :\\; g \\geq\\lambda \\}| \\leq e^{-\n\\lambda^{2}/(2^{d}\\|S(g)\\|_{\\infty}^{2})}, \\] where $S(g)$ denotes the\nclassical dyadic square function in $[0,1)^{d}$. The estimate is sharp when\ndimension $d$ tends to infinity in the sense that the constant $2^{d}$ in the\ndenominator cannot be replaced by $C2^{d}$ with $0<C<1$ independent of $d$ when\n$d \\to \\infty$.\nFor $d=1$ this is a classical result of Chang--Wilson--Wolff [4]; however, in\nthe case $d>1$ they work with a special square function $S_\\infty$, and their\nresult does not imply the estimates for the classical square function.\nUsing good $\\lambda$ inequalities technique we then obtain unweighted and\nweighted $L^p$ lower bounds for $S$; to get the corresponding good $\\lambda$\ninequalities we need to modify the classical construction.\nWe also show how to obtain our superexponential distribution inequality\n(although with worse constants) from the weighted $L^2$ lower bounds for $S$,\nobtained in [5].\n",
"title": "Superexponential estimates and weighted lower bounds for the square function"
}
| null | null | null | null | true | null |
18067
| null |
Default
| null | null |
null |
{
"abstract": " Since the beginning of the new millennium, stock markets went through every\nstate from long-time troughs, trade suspensions to all-time highs. The\nliterature on asset pricing hence assumes random processes to be underlying the\nmovement of stock returns. Observed procyclicality and time-varying correlation\nof stock returns tried to give the apparently random behavior some sort of\nstructure. However, common misperceptions about the co-movement of asset prices\nin the years preceding the \\emph{Great Recession} and the \\emph{Global\nCommodity Crisis}, is said to have even fueled the crisis' economic impact.\nHere we show how a varying macroeconomic environment influences stocks'\nclustering into communities. From a sample of 296 stocks of the S\\&P 500 index,\ndistinct periods in between 2004 and 2011 are used to develop networks of\nstocks. The Minimal Spanning Tree analysis of those time-varying networks of\nstocks demonstrates that the crises of 2007-2008 and 2010-2011 drove the market\nto clustered community structures in both periods, helping to restore the stock\nmarket's ceased order of the pre-crises era. However, a comparison of the\nemergent clusters with the \\textit{General Industry Classification Standard}\nconveys the impression that industry sectors do not play a major role in that\norder.\n",
"title": "Reframing the S\\&P500 Network of Stocks along the \\nth{21} Century"
}
| null | null | null | null | true | null |
18068
| null |
Default
| null | null |
null |
{
"abstract": " The superconducting energy gap in $\\rm DyNi_2B_2C$ has been investigated\nusing a point-contact technique based on the Andreev reflection from a normal\n(N)-superconductor (S) boundary, where N is Ag. The observed differential\nresistance $dV/dI$ is well described by the Blonder-Tinkham-Klapwijk (BTK)\ntheory based on the BSC density of states with zero broadening parameter.\nTypically, the intensity of the gap structure amounts to several percentage of\nthe normal state resistance, which is an order of magnitude less than predicted\nby the theory. For $\\rm DyNi_2B_2C$ with $T_c<T_N$ (the Neel temperature), we\nfound gap values satisfying the ratio of $2\\Delta_0/k_BT_c=3.63\\pm 0.05$\nsimilar to other superconducting nickel-borocarbides, both nonmagnetic and\nmagnetic with $T_c\\geq T_N$. The superconducting gap nonlinearity is\nsuperimposed on the antiferromagnetic structure in $dV/dI(V)$ which is\nsuppressed at the magnetic field of the order of 3T applied nominally in the\n$ab$-plane and temperature $\\geq 11~K$. The observed superconducting properties\ndepend on the exact composition and structure at the surface of the crystal.\n",
"title": "Point-contact spectroscopy of superconducting energy gap in $\\rm DyNi_2B_2C$"
}
| null | null | null | null | true | null |
18069
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we study the problem of finding a small safe set $S$ in a graph\n$G$, i.e. a non-empty set of vertices such that no connected component of\n$G[S]$ is adjacent to a larger component in $G - S$. We enhance our\nunderstanding of the problem from the viewpoint of parameterized complexity by\nshowing that (1) the problem is W[2]-hard when parameterized by the pathwidth\n$pw$ and cannot be solved in time $n^{o(pw)}$ unless the ETH is false, (2) it\nadmits no polynomial kernel parameterized by the vertex cover number $vc$\nunless $\\mathrm{PH} = \\Sigma^{\\mathrm{p}}_{3}$, but (3) it is fixed-parameter\ntractable (FPT) when parameterized by the neighborhood diversity $nd$, and (4)\nit can be solved in time $n^{f(cw)}$ for some double exponential function $f$\nwhere $cw$ is the clique-width. We also present (5) a faster FPT algorithm when\nparameterized by solution size.\n",
"title": "Parameterized Complexity of Safe Set"
}
| null | null | null | null | true | null |
18070
| null |
Default
| null | null |
null |
{
"abstract": " Although there is ample work in the literature dealing with skewness in the\nmultivariate setting, there is a relative paucity of work in the matrix variate\nparadigm. Such work is, for example, useful for modelling three-way data. A\nmatrix variate skew-t distribution is derived based on a mean-variance matrix\nnormal mixture. An expectation-conditional maximization algorithm is developed\nfor parameter estimation. Simulated data are used for illustration.\n",
"title": "A Matrix Variate Skew-t Distribution"
}
| null | null | null | null | true | null |
18071
| null |
Default
| null | null |
null |
{
"abstract": " Using age of information as the freshness metric, we examine a multicast\nnetwork in which real-time status updates are generated by the source and sent\nto a group of $n$ interested receivers. We show that in order to keep the\ninformation freshness at each receiver, the source should terminate the\ntransmission of the current update and start sending a new update packet as\nsoon as it receives the acknowledgements back from any $k$ out of $n$ nodes. As\nthe source stopping threshold $k$ increases, a node is more likely to get the\nlatest generated update, but the age of the most recent update is more likely\nto become outdated. We derive the age minimized stopping threshold $k$ that\nbalances the likelihood of getting the latest update and the freshness of the\nlatest update for shifted exponential link delay. Through numerical evaluations\nfor different stopping strategies, we find that waiting for the\nacknowledgements from the earliest $k$ out of $n$ nodes leads to lower average\nage than waiting for a pre-selected group of $k$ nodes. We also observe that a\nproperly chosen threshold $k$ can prevent information staleness for increasing\nnumber of nodes $n$ in the multicast network.\n",
"title": "Status Updates Through Multicast Networks"
}
| null | null | null | null | true | null |
18072
| null |
Default
| null | null |
null |
{
"abstract": " We applied pre-defined kernels also known as filters or masks developed for\nimage processing to convolution neural network. Instead of letting neural\nnetworks find its own kernels, we used 41 different general-purpose kernels of\nblurring, edge detecting, sharpening, discrete cosine transformation, etc. for\nthe first layer of the convolution neural networks. This architecture, thus\nnamed as general filter convolutional neural network (GFNN), can reduce\ntraining time by 30% with a better accuracy compared to the regular\nconvolutional neural network (CNN). GFNN also can be trained to achieve 90%\naccuracy with only 500 samples. Furthermore, even though these kernels are not\nspecialized for the MNIST dataset, we achieved 99.56% accuracy without ensemble\nnor any other special algorithms.\n",
"title": "Extension of Convolutional Neural Network with General Image Processing Kernels"
}
| null | null | null | null | true | null |
18073
| null |
Default
| null | null |
null |
{
"abstract": " This paper proposes a neural network architecture and training scheme to\nlearn the start and end time of sound events (strong labels) in an audio\nrecording given just the list of sound events existing in the audio without\ntime information (weak labels). We achieve this by using a stacked\nconvolutional and recurrent neural network with two prediction layers in\nsequence one for the strong followed by the weak label. The network is trained\nusing frame-wise log mel-band energy as the input audio feature, and weak\nlabels provided in the dataset as labels for the weak label prediction layer.\nStrong labels are generated by replicating the weak labels as many number of\ntimes as the frames in the input audio feature, and used for strong label layer\nduring training. We propose to control what the network learns from the weak\nand strong labels by different weighting for the loss computed in the two\nprediction layers. The proposed method is evaluated on a publicly available\ndataset of 155 hours with 17 sound event classes. The method achieves the best\nerror rate of 0.84 for strong labels and F-score of 43.3% for weak labels on\nthe unseen test split.\n",
"title": "Sound event detection using weakly labeled dataset with stacked convolutional and recurrent neural network"
}
| null | null | null | null | true | null |
18074
| null |
Default
| null | null |
null |
{
"abstract": " Let $(\\mathcal{K} ,\\subseteq )$ be a universal class with\n$LS(\\mathcal{K})=\\lambda$ categorical in regular $\\kappa >\\lambda^+$ with\narbitrarily large models, and let $\\mathcal{K}^*$ be the class of all\n$\\mathcal{A}\\in\\mathcal{K}_{>\\lambda}$ for which there is $\\mathcal{B} \\in\n\\mathcal{K}_{\\ge\\kappa}$ such that $\\mathcal{A}\\subseteq\\mathcal{B}$. We prove\nthat $\\mathcal{K}^*$ is categorical in every $\\xi >\\lambda^+$,\n$\\mathcal{K}_{\\ge\\beth_{(2^{\\lambda^+})^+}} \\subseteq \\mathcal{K}^{*}$, and the\nmodels of $\\mathcal{K}^*_{>\\lambda^+}$ are essentially vector spaces (or\ntrivial i.e. disintegrated).\n",
"title": "Categoricity and Universal Classes"
}
| null | null | null | null | true | null |
18075
| null |
Default
| null | null |
null |
{
"abstract": " The super-Neptune exoplanet WASP-107b is an exciting target for atmosphere\ncharacterization. It has an unusually large atmospheric scale height and a\nsmall, bright host star, raising the possibility of precise constraints on its\ncurrent nature and formation history. We report the first atmospheric study of\nWASP-107b, a Hubble Space Telescope measurement of its near-infrared\ntransmission spectrum. We determined the planet's composition with two\ntechniques: atmospheric retrieval based on the transmission spectrum and\ninterior structure modeling based on the observed mass and radius. The interior\nstructure models set a $3\\,\\sigma$ upper limit on the atmospheric metallicity\nof $30\\times$ solar. The transmission spectrum shows strong evidence for water\nabsorption ($6.5\\,\\sigma$ confidence), and the retrieved water abundance is\nconsistent with expectations for a solar abundance pattern. The inferred\ncarbon-to-oxygen ratio is subsolar at $2.7\\,\\sigma$ confidence, which we\nattribute to possible methane depletion in the atmosphere. The spectral\nfeatures are smaller than predicted for a cloud-free composition, crossing less\nthan one scale height. A thick condensate layer at high altitudes (0.1 - 3\nmbar) is needed to match the observations. We find that physically motivated\ncloud models with moderate sedimentation efficiency ($f_\\mathrm{sed} = 0.3$) or\nhazes with a particle size of 0.3 $\\mu$m reproduce the observed spectral\nfeature amplitude. Taken together, these findings serve as an illustration of\nthe diversity and complexity of exoplanet atmospheres. The community can look\nforward to more such results with the high precision and wide spectral coverage\nafforded by future observing facilities.\n",
"title": "Water, High-Altitude Condensates, and Possible Methane Depletion in the Atmosphere of the Warm Super-Neptune WASP-107b"
}
| null | null | null | null | true | null |
18076
| null |
Default
| null | null |
null |
{
"abstract": " Logical models have been successfully used to describe regulatory and\nsignaling networks without requiring quantitative data. However, existing data\nis insufficient to adequately define a unique model, rendering the\nparametrization of a given model a difficult task.\nHere, we focus on the characterization of the set of Boolean functions\ncompatible with a given regulatory structure, i.e. the set of all monotone\nnondegenerate Boolean functions. We then propose an original set of rules to\nlocally explore the direct neighboring functions of any function in this set,\nwithout explicitly generating the whole set. Also, we provide relationships\nbetween the regulatory functions and their corresponding dynamics.\nFinally, we illustrate the usefulness of this approach by revisiting\nProbabilistic Boolean Networks with the model of T helper cell differentiation\nfrom Mendoza & Xenarios.\n",
"title": "Partial Order on the set of Boolean Regulatory Functions"
}
| null | null | null | null | true | null |
18077
| null |
Default
| null | null |
null |
{
"abstract": " In this work, we present scalable balancing domain decomposition by\nconstraints methods for linear systems arising from arbitrary order edge finite\nelement discretizations of multi-material and heterogeneous 3D problems. In\norder to enforce the continuity across subdomains of the method, we use a\npartition of the interface objects (edges and faces) into sub-objects\ndetermined by the variation of the physical coefficients of the problem. For\nmulti-material problems, a constant coefficient condition is enough to define\nthis sub-partition of the objects. For arbitrarily heterogeneous problems, a\nrelaxed version of the method is defined, where we only require that the\nmaximal contrast of the physical coefficient in each object is smaller than a\npredefined threshold. Besides, the addition of perturbation terms to the\npreconditioner is empirically shown to be effective in order to deal with the\ncase where the two coefficients of the model problem jump simultaneously across\nthe interface. The new method, in contrast to existing approaches for problems\nin curl-conforming spaces, preserves the simplicity of the original\npreconditioner, i.e., no spectral information is required, whilst providing\nrobustness with regard to coefficient jumps and heterogeneous materials. A\ndetailed set of numerical experiments, which includes the application of the\npreconditioner to 3D realistic cases, shows excellent weak scalability\nproperties of the implementation of the proposed algorithms.\n",
"title": "Scalable solvers for complex electromagnetics problems"
}
| null | null | null | null | true | null |
18078
| null |
Default
| null | null |
null |
{
"abstract": " A flux-splitting method is proposed for the hyperbolic-equation system (HES)\nof magnetized electron fluids in quasi-neutral plasmas. The numerical fluxes\nare split into four categories, which are computed by using an upwind method\nwhich incorporates a flux-vector splitting (FVS) and advection upstream\nsplitting method (AUSM). The method is applied to a test calculation condition\nof uniformly distributed and angled magnetic lines of force. All of the\npseudo-time advancement terms converge monotonically and the conservation laws\nare strictly satisfied in the steady state. The calculation results are\ncompared with those computed by using the elliptic-parabolic-equation system\n(EPES) approach using a magnetic-field-aligned mesh (MFAM). Both qualitative\nand quantitative comparisons yield good agreements of results, indicating that\nthe HES approach with the flux-splitting method attains a high computational\naccuracy.\n",
"title": "A flux-splitting method for hyperbolic-equation system of magnetized electron fluids in quasi-neutral plasmas"
}
| null | null | null | null | true | null |
18079
| null |
Default
| null | null |
null |
{
"abstract": " We address unsupervised optical flow estimation for ego-centric motion. We\nargue that optical flow can be cast as a geometrical warping between two\nsuccessive video frames and devise a deep architecture to estimate such\ntransformation in two stages. First, a dense pixel-level flow is computed with\na geometric prior imposing strong spatial constraints. Such prior is typical of\ndriving scenes, where the point of view is coherent with the vehicle motion. We\nshow how such global transformation can be approximated with an homography and\nhow spatial transformer layers can be employed to compute the flow field\nimplied by such transformation. The second stage then refines the prediction\nfeeding a second deeper network. A final reconstruction loss compares the\nwarping of frame X(t) with the subsequent frame X(t+1) and guides both\nestimates. The model, which we named TransFlow, performs favorably compared to\nother unsupervised algorithms, and shows better generalization compared to\nsupervised methods with a 3x reduction in error on unseen data.\n",
"title": "TransFlow: Unsupervised Motion Flow by Joint Geometric and Pixel-level Estimation"
}
| null | null | null | null | true | null |
18080
| null |
Default
| null | null |
null |
{
"abstract": " Autonomous driving requires 3D perception of vehicles and other objects in\nthe in environment. Much of the current methods support 2D vehicle detection.\nThis paper proposes a flexible pipeline to adopt any 2D detection network and\nfuse it with a 3D point cloud to generate 3D information with minimum changes\nof the 2D detection networks. To identify the 3D box, an effective model\nfitting algorithm is developed based on generalised car models and score maps.\nA two-stage convolutional neural network (CNN) is proposed to refine the\ndetected 3D box. This pipeline is tested on the KITTI dataset using two\ndifferent 2D detection networks. The 3D detection results based on these two\nnetworks are similar, demonstrating the flexibility of the proposed pipeline.\nThe results rank second among the 3D detection algorithms, indicating its\ncompetencies in 3D detection.\n",
"title": "A General Pipeline for 3D Detection of Vehicles"
}
| null | null | null | null | true | null |
18081
| null |
Default
| null | null |
null |
{
"abstract": " We present a search for CII emission over cosmological scales at\nhigh-redshifts. The CII line is a prime candidate to be a tracer of star\nformation over large-scale structure since it is one of the brightest emission\nlines from galaxies. Redshifted CII emission appears in the submillimeter\nregime, meaning it could potentially be present in the higher frequency\nintensity data from the Planck satellite used to measure the cosmic infrared\nbackground (CIB). We search for CII emission over redshifts z=2-3.2 in the\nPlanck 545 GHz intensity map by cross-correlating the 3 highest frequency\nPlanck maps with spectroscopic quasars and CMASS galaxies from the Sloan\nDigital Sky Survey III (SDSS-III), which we then use to jointly fit for CII\nintensity, CIB parameters, and thermal Sunyaev-Zeldovich (SZ) emission. We\nreport a measurement of an anomalous emission\n$\\mathrm{I_\\nu}=6.6^{+5.0}_{-4.8}\\times10^4$ Jy/sr at 95% confidence, which\ncould be explained by CII emission, favoring collisional excitation models of\nCII emission that tend to be more optimistic than models based on CII\nluminosity scaling relations from local measurements; however, a comparison of\nBayesian information criteria reveal that this model and the CIB & SZ only\nmodel are equally plausible. Thus, more sensitive measurements will be needed\nto confirm the existence of large-scale CII emission at high redshifts.\nFinally, we forecast that intensity maps from Planck cross-correlated with\nquasars from the Dark Energy Spectroscopic Instrument (DESI) would increase our\nsensitivity to CII emission by a factor of 5, while the proposed Primordial\nInflation Explorer (PIXIE) could increase the sensitivity further.\n",
"title": "Search for CII Emission on Cosmological Scales at Redshift Z~2.6"
}
| null | null | null | null | true | null |
18082
| null |
Default
| null | null |
null |
{
"abstract": " We complete the proof of the Generalized Smale Conjecture, apart from the\ncase of $RP^3$, and give a new proof of Gabai's theorem for hyperbolic\n3-manifolds. We use an approach based on Ricci flow through singularities,\nwhich applies uniformly to spherical space forms other than $S^3$ and $RP^3$\nand hyperbolic manifolds, to prove that the moduli space of metrics of constant\nsectional curvature is contractible. As a corollary, for such a 3-manifold $X$,\nthe inclusion $\\text{Isom} (X,g)\\to \\text{Diff}(X)$ is a homotopy equivalence\nfor any Riemannian metric $g$ of constant sectional curvature.\n",
"title": "Ricci flow and diffeomorphism groups of 3-manifolds"
}
| null | null | null | null | true | null |
18083
| null |
Default
| null | null |
null |
{
"abstract": " Reinforcement learning (RL) algorithms involve the deep nesting of highly\nirregular computation patterns, each of which typically exhibits opportunities\nfor distributed computation. We argue for distributing RL components in a\ncomposable way by adapting algorithms for top-down hierarchical control,\nthereby encapsulating parallelism and resource requirements within\nshort-running compute tasks. We demonstrate the benefits of this principle\nthrough RLlib: a library that provides scalable software primitives for RL.\nThese primitives enable a broad range of algorithms to be implemented with high\nperformance, scalability, and substantial code reuse. RLlib is available at\nthis https URL.\n",
"title": "RLlib: Abstractions for Distributed Reinforcement Learning"
}
| null | null | null | null | true | null |
18084
| null |
Default
| null | null |
null |
{
"abstract": " A multiscale analysis of 1D stochastic bistable reaction-diffusion equations\nwith additive noise is carried out w.r.t. travelling waves within the\nvariational approach to stochastic partial differential equations. It is shown\nwith explicit error estimates on appropriate function spaces that up to lower\norder w.r.t. the noise amplitude, the solution can be decomposed into the\northogonal sum of a travelling wave moving with random speed and into Gaussian\nfluctuations. A stochastic differential equation describing the speed of the\ntravelling wave and a linear stochastic partial differential equation\ndescribing the fluctuations are derived in terms of the coefficients. Our\nresults extend corresponding results obtained for stochastic neural field\nequations to the present class of stochastic dynamics.\n",
"title": "A Multiscale-Analysis of Stochastic Bistable Reaction-Diffusion Equations"
}
| null | null | null | null | true | null |
18085
| null |
Default
| null | null |
null |
{
"abstract": " Interpretability has emerged as a crucial aspect of machine learning, aimed\nat providing insights into the working of complex neural networks. However,\nexisting solutions vary vastly based on the nature of the interpretability\ntask, with each use case requiring substantial time and effort. This paper\nintroduces MARGIN, a simple yet general approach to address a large set of\ninterpretability tasks ranging from identifying prototypes to explaining image\npredictions. MARGIN exploits ideas rooted in graph signal analysis to determine\ninfluential nodes in a graph, which are defined as those nodes that maximally\ndescribe a function defined on the graph. By carefully defining task-specific\ngraphs and functions, we demonstrate that MARGIN outperforms existing\napproaches in a number of disparate interpretability challenges.\n",
"title": "MARGIN: Uncovering Deep Neural Networks using Graph Signal Analysis"
}
| null | null | null | null | true | null |
18086
| null |
Default
| null | null |
null |
{
"abstract": " We propose a new technique, Singular Vector Canonical Correlation Analysis\n(SVCCA), a tool for quickly comparing two representations in a way that is both\ninvariant to affine transform (allowing comparison between different layers and\nnetworks) and fast to compute (allowing more comparisons to be calculated than\nwith previous methods). We deploy this tool to measure the intrinsic\ndimensionality of layers, showing in some cases needless over-parameterization;\nto probe learning dynamics throughout training, finding that networks converge\nto final representations from the bottom up; to show where class-specific\ninformation in networks is formed; and to suggest new training regimes that\nsimultaneously save computation and overfit less. Code:\nthis https URL\n",
"title": "SVCCA: Singular Vector Canonical Correlation Analysis for Deep Learning Dynamics and Interpretability"
}
| null | null | null | null | true | null |
18087
| null |
Default
| null | null |
null |
{
"abstract": " This paper establishes a general equivalence between discrete choice and\nrational inattention models. Matejka and McKay (2015, AER) showed that when\ninformation costs are modelled using the Shannon entropy function, the\nresulting choice probabilities in the rational inattention model take the\nmultinomial logit form. By exploiting convex-analytic properties of the\ndiscrete choice model, we show that when information costs are modelled using a\nclass of generalized entropy functions, the choice probabilities in any\nrational inattention model are observationally equivalent to some additive\nrandom utility discrete choice model and vice versa. Thus any additive random\nutility model can be given an interpretation in terms of boundedly rational\nbehavior. This includes empirically relevant specifications such as the probit\nand nested logit models.\n",
"title": "Discrete Choice and Rational Inattention: a General Equivalence Result"
}
| null | null | null | null | true | null |
18088
| null |
Default
| null | null |
null |
{
"abstract": " We prove a Hardy inequality for ultraspherical expansions by using a proper\nground state representation. From this result we deduce some uncertainty\nprinciples for this kind of expansions. Our result also implies a Hardy\ninequality on spheres with a potential having a double singularity.\n",
"title": "A Hardy inequality for ultraspherical expansions with an application to the sphere"
}
| null | null | null | null | true | null |
18089
| null |
Default
| null | null |
null |
{
"abstract": " Bayesian Neural Networks (BNNs) have recently received increasing attention\nfor their ability to provide well-calibrated posterior uncertainties. However,\nmodel selection---even choosing the number of nodes---remains an open question.\nRecent work has proposed the use of a horseshoe prior over node pre-activations\nof a Bayesian neural network, which effectively turns off nodes that do not\nhelp explain the data. In this work, we propose several modeling and inference\nadvances that consistently improve the compactness of the model learned while\nmaintaining predictive performance, especially in smaller-sample settings\nincluding reinforcement learning.\n",
"title": "Structured Variational Learning of Bayesian Neural Networks with Horseshoe Priors"
}
| null | null | null | null | true | null |
18090
| null |
Default
| null | null |
null |
{
"abstract": " In several realistic situations, an interactive learning agent can practice\nand refine its strategy before going on to be evaluated. For instance, consider\na student preparing for a series of tests. She would typically take a few\npractice tests to know which areas she needs to improve upon. Based of the\nscores she obtains in these practice tests, she would formulate a strategy for\nmaximizing her scores in the actual tests. We treat this scenario in the\ncontext of an agent exploring a fixed-horizon episodic Markov Decision Process\n(MDP), where the agent can practice on the MDP for some number of episodes (not\nnecessarily known in advance) before starting to incur regret for its actions.\nDuring practice, the agent's goal must be to maximize the probability of\nfollowing an optimal policy. This is akin to the problem of Pure Exploration\n(PE). We extend the PE problem of Multi Armed Bandits (MAB) to MDPs and propose\na Bayesian algorithm called Posterior Sampling for Pure Exploration (PSPE),\nwhich is similar to its bandit counterpart. We show that the Bayesian simple\nregret converges at an optimal exponential rate when using PSPE.\nWhen the agent starts being evaluated, its goal would be to minimize the\ncumulative regret incurred. This is akin to the problem of Reinforcement\nLearning (RL). The agent uses the Posterior Sampling for Reinforcement Learning\nalgorithm (PSRL) initialized with the posteriors of the practice phase. We\nhypothesize that this PSPE + PSRL combination is an optimal strategy for\nminimizing regret in RL problems with an initial practice phase. We show\nempirical results which prove that having a lower simple regret at the end of\nthe practice phase results in having lower cumulative regret during evaluation.\n",
"title": "Efficient Reinforcement Learning via Initial Pure Exploration"
}
| null | null | null | null | true | null |
18091
| null |
Default
| null | null |
null |
{
"abstract": " This paper compares the results of applying a recently developed method of\nstochastic uncertainty quantification designed for fluid dynamics to the\nBorn-Infeld model of nonlinear electromagnetism. The similarities in the\nresults are striking. Namely, the introduction of Stratonovich cylindrical\nnoise into each of their Hamiltonian formulations introduces stochastic Lie\ntransport into their dynamics in the same form for both theories. Moreover, the\nresulting stochastic partial differential equations (SPDE) retain their\nunperturbed form, except for an additional term representing induced Lie\ntransport by the set of divergence-free vector fields associated with the\nspatial correlations of the cylindrical noise. The explanation for this\nremarkable similarity lies in the method of construction of the Hamiltonian for\nthe Stratonovich stochastic contribution to the motion in both cases; which is\ndone via pairing spatial correlation eigenvectors for cylindrical noise with\nthe momentum map for the deterministic motion. This momentum map is responsible\nfor the well-known analogy between hydrodynamics and electromagnetism. The\nmomentum map for the Maxwell and Born-Infeld theories of electromagnetism\ntreated here is the 1-form density known as the Poynting vector. Two Appendices\ntreat the Hamiltonian structures underlying these results.\n",
"title": "Stochastic Evolution of Augmented Born--Infeld Equations"
}
| null | null | null | null | true | null |
18092
| null |
Default
| null | null |
null |
{
"abstract": " Acoustical radiation force (ARF) induced by a single Bessel beam with\narbitrary order and location on a nonspherical shape is studied with the\nemphasis on the physical mechanism and parameter conditions of negative\n(pulling) forces. Numerical experiments are conducted to verify the T-matrix\nmethod (TMM) for axial ARFs. This study may guide the experimental set-up to\nfind negative axial ARF quickly and effectively based on the predicted\nparameters with TMM, and could be extended for lateral forces. The present work\ncould help to design acoustic tweezers numerical toolbox, which provides an\nalternate to the optic tweezers.\n",
"title": "T-matrix evaluation of acoustic radiation forces on nonspherical objects in Bessel beams"
}
| null | null | null | null | true | null |
18093
| null |
Default
| null | null |
null |
{
"abstract": " We construct two infinite series of Moufang loops of exponent $3$ whose\ncommutative center (i.e. the set of elements that commute with all elements of\nthe loop) is not a normal subloop. In particular, we obtain examples of such\nloops of orders $3^8$ and $3^{11}$ one of which can be defined as the Moufang\ntriplication of the free Burnside group $B(3,3)$.\n",
"title": "On the commutative center of Moufang loops"
}
| null | null |
[
"Mathematics"
] | null | true | null |
18094
| null |
Validated
| null | null |
null |
{
"abstract": " In the paper \"Formality conjecture\" (1996) Kontsevich designed a universal\nflow $\\dot{\\mathcal{P}}=\\mathcal{Q}_{a:b}(\\mathcal{P})=a\\Gamma_{1}+b\\Gamma_{2}$\non the spaces of Poisson structures $\\mathcal{P}$ on all affine manifolds of\ndimension $n \\geqslant 2$. We prove a claim from $\\textit{loc. cit.}$ stating\nthat if $n=2$, the flow $\\mathcal{Q}_{1:0}=\\Gamma_{1}(\\mathcal{P})$ is\nPoisson-cohomology trivial: $\\Gamma_{1}(\\mathcal{P})$ is the Schouten bracket\nof $\\mathcal{P}$ with $\\mathcal{X}$, for some vector field $\\mathcal{X}$; we\nexamine the structure of the space of solutions $\\mathcal{X}$. Both the\nconstruction of differential polynomials $\\Gamma_{1}(\\mathcal{P})$ and\n$\\Gamma_{2}(\\mathcal{P})$ and the technique to study them remain valid in\nhigher dimensions $n \\geqslant 3$, but neither the trivializing vector field\n$\\mathcal{X}$ nor the setting $b:=0$ survive at $n\\geqslant 3$, where the\nbalance is $a:b=1:6$.\n",
"title": "The Kontsevich tetrahedral flow in 2D: a toy model"
}
| null | null | null | null | true | null |
18095
| null |
Default
| null | null |
null |
{
"abstract": " Many similarity-based clustering methods work in two separate steps including\nsimilarity matrix computation and subsequent spectral clustering. However,\nsimilarity measurement is challenging because it is usually impacted by many\nfactors, e.g., the choice of similarity metric, neighborhood size, scale of\ndata, noise and outliers. Thus the learned similarity matrix is often not\nsuitable, let alone optimal, for the subsequent clustering. In addition,\nnonlinear similarity often exists in many real world data which, however, has\nnot been effectively considered by most existing methods. To tackle these two\nchallenges, we propose a model to simultaneously learn cluster indicator matrix\nand similarity information in kernel spaces in a principled way. We show\ntheoretical relationships to kernel k-means, k-means, and spectral clustering\nmethods. Then, to address the practical issue of how to select the most\nsuitable kernel for a particular clustering task, we further extend our model\nwith a multiple kernel learning ability. With this joint model, we can\nautomatically accomplish three subtasks of finding the best cluster indicator\nmatrix, the most accurate similarity relations and the optimal combination of\nmultiple kernels. By leveraging the interactions between these three subtasks\nin a joint framework, each subtask can be iteratively boosted by using the\nresults of the others towards an overall optimal solution. Extensive\nexperiments are performed to demonstrate the effectiveness of our method.\n",
"title": "Twin Learning for Similarity and Clustering: A Unified Kernel Approach"
}
| null | null | null | null | true | null |
18096
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we apply shrinkage strategies to estimate regression\ncoefficients efficiently for the high-dimensional multiple regression model,\nwhere the number of samples is smaller than the number of predictors. We assume\nin the sparse linear model some of the predictors have very weak influence on\nthe response of interest. We propose to shrink estimators more than usual.\nSpecifically, we use integrated estimation strategies in sub and full models\nand shrink the integrated estimators by incorporating a bounded measurable\nfunction of some weights. The exhibited double shrunken estimators improve the\nprediction performance of sub models significantly selected from existing\nLasso-type variable selection methods. Monte Carlo simulation studies as well\nas real examples of eye data and Riboavin data confirm the superior performance\nof the estimators in the high-dimensional regression model.\n",
"title": "Big Data Analysis Using Shrinkage Strategies"
}
| null | null | null | null | true | null |
18097
| null |
Default
| null | null |
null |
{
"abstract": " Let M be a transitive model of set theory. There is a canonical\ninterpretation functor between the category of regular Hausdorff, continuous\nopen images of Cech-complete spaces of M and the same category in V, preserving\nmany concepts of topology, functional analysis, and dynamics. The functor can\nbe further canonically extended to the category of Borel subspaces. This\ngreatly simplifies and extends similar results of Fremlin.\n",
"title": "Interpreter fr topologists"
}
| null | null | null | null | true | null |
18098
| null |
Default
| null | null |
null |
{
"abstract": " We propose a parallel-data-free voice-conversion (VC) method that can learn a\nmapping from source to target speech without relying on parallel data. The\nproposed method is general purpose, high quality, and parallel-data free and\nworks without any extra data, modules, or alignment procedure. It also avoids\nover-smoothing, which occurs in many conventional statistical model-based VC\nmethods. Our method, called CycleGAN-VC, uses a cycle-consistent adversarial\nnetwork (CycleGAN) with gated convolutional neural networks (CNNs) and an\nidentity-mapping loss. A CycleGAN learns forward and inverse mappings\nsimultaneously using adversarial and cycle-consistency losses. This makes it\npossible to find an optimal pseudo pair from unpaired data. Furthermore, the\nadversarial loss contributes to reducing over-smoothing of the converted\nfeature sequence. We configure a CycleGAN with gated CNNs and train it with an\nidentity-mapping loss. This allows the mapping function to capture sequential\nand hierarchical structures while preserving linguistic information. We\nevaluated our method on a parallel-data-free VC task. An objective evaluation\nshowed that the converted feature sequence was near natural in terms of global\nvariance and modulation spectra. A subjective evaluation showed that the\nquality of the converted speech was comparable to that obtained with a Gaussian\nmixture model-based method under advantageous conditions with parallel and\ntwice the amount of data.\n",
"title": "Parallel-Data-Free Voice Conversion Using Cycle-Consistent Adversarial Networks"
}
| null | null | null | null | true | null |
18099
| null |
Default
| null | null |
null |
{
"abstract": " Least-squares models such as linear regression and Linear Discriminant\nAnalysis (LDA) are amongst the most popular statistical learning techniques.\nHowever, since their computation time increases cubically with the number of\nfeatures, they are inefficient in high-dimensional neuroimaging datasets.\nFortunately, for k-fold cross-validation, an analytical approach has been\ndeveloped that yields the exact cross-validated predictions in least-squares\nmodels without explicitly training the model. Its computation time grows with\nthe number of test samples. Here, this approach is systematically investigated\nin the context of cross-validation and permutation testing. LDA is used\nexemplarily but results hold for all other least-squares methods. Furthermore,\na non-trivial extension to multi-class LDA is formally derived. The analytical\napproach is evaluated using complexity calculations, simulations, and\npermutation testing of an EEG/MEG dataset. Depending on the ratio between\nfeatures and samples, the analytical approach is up to 10,000x faster than the\nstandard approach (retraining the model on each training set). This allows for\na fast cross-validation of least-squares models and multi-class LDA in\nhigh-dimensional data, with obvious applications in multi-dimensional datasets,\nRepresentational Similarity Analysis, and permutation testing.\n",
"title": "Cross-validation in high-dimensional spaces: a lifeline for least-squares models and multi-class LDA"
}
| null | null | null | null | true | null |
18100
| null |
Default
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.