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": " Context. The 4th release of the SDSS Moving Object Catalog (SDSSMOC) is\npresently the largest photometric dataset of asteroids. Up to this point, the\nrelease of large asteroid datasets has always been followed by a redefinition\nof asteroid taxonomy. In the years that followed the release of the first\nSDSSMOC, several classification schemes using its data were proposed, all using\nthe taxonomic classes from previous taxonomies. However, no successful attempt\nhas been made to derive a new taxonomic system directly from the SDSS dataset.\nAims. The scope of the work is to propose a different interpretation scheme for\ngauging u0g0r0i0z0 asteroid observations based on the continuity of spectral\nfeatures. The scheme is integrated into previous taxonomic labeling, but is not\ndependent on them. Methods. We analyzed the behavior of asteroid sampling\nthrough principal components analysis to understand the role of uncertainties\nin the SDSSMOC. We identified that asteroids in this space follow two separate\nlinear trends using reflectances in the visible, which is characteristic of\ntheir spectrophotometric features. Results. Introducing taxonomic classes, we\nare able to interpret both trends as representative of featured and featureless\nspectra. The evolution within the trend is connected mainly to the band depth\nfor featured asteroids and to the spectral slope for featureless ones. We\ndefined a different taxonomic system that allowed us to only classify asteroids\nby two labels. Conclusions. We have classified 69% of all SDSSMOC sample, which\nis a robustness higher than reached by previous SDSS classifications.\nFurthermore, as an example, we present the behavior of asteroid (5129) Groom,\nwhose taxonomic labeling changes according to one of the trends owing to phase\nreddening. Now, such behavior can be characterized by the variation of one\nsingle parameter, its position in the trend.\n",
"title": "Characterizing spectral continuity in SDSS u'g'r'i'z' asteroid photometry"
}
| null | null | null | null | true | null |
15601
| null |
Default
| null | null |
null |
{
"abstract": " The critical behavior of the random transverse-field Ising model in finite\ndimensional lattices is governed by infinite disorder fixed points, several\nproperties of which have already been calculated by the use of the strong\ndisorder renormalization group (SDRG) method. Here we extend these studies and\ncalculate the connected transverse-spin correlation function by a numerical\nimplementation of the SDRG method in $d=1,2$ and $3$ dimensions. At the\ncritical point an algebraic decay of the form $\\sim r^{-\\eta_t}$ is found, with\na decay exponent being approximately $\\eta_t \\approx 2+2d$. In $d=1$ the\nresults are related to dimer-dimer correlations in the random AF XX-chain and\nhave been tested by numerical calculations using free-fermionic techniques.\n",
"title": "Transverse-spin correlations of the random transverse-field Ising model"
}
| null | null | null | null | true | null |
15602
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we extend the known methodology for fitting stable\ndistributions to the multivariate case and apply the suggested method to the\nmodelling of daily cryptocurrency-return data. The investigated time period is\ncut into 10 non-overlapping sections, thus the changes can also be observed. We\napply bootstrap tests for checking the models and compare our approach to the\nmore traditional extreme-value and copula models.\n",
"title": "Multivariate stable distributions and their applications for modelling cryptocurrency-returns"
}
| null | null | null | null | true | null |
15603
| null |
Default
| null | null |
null |
{
"abstract": " User participation in online communities is driven by the intertwinement of\nthe social network structure with the crowd-generated content that flows along\nits links. These aspects are rarely explored jointly and at scale. By looking\nat how users generate and access pictures of varying beauty on Flickr, we\ninvestigate how the production of quality impacts the dynamics of online social\nsystems. We develop a deep learning computer vision model to score images\naccording to their aesthetic value and we validate its output through\ncrowdsourcing. By applying it to over 15B Flickr photos, we study for the first\ntime how image beauty is distributed over a large-scale social system.\nBeautiful images are evenly distributed in the network, although only a small\ncore of people get social recognition for them. To study the impact of exposure\nto quality on user engagement, we set up matching experiments aimed at\ndetecting causality from observational data. Exposure to beauty is\ndouble-edged: following people who produce high-quality content increases one's\nprobability of uploading better photos; however, an excessive imbalance between\nthe quality generated by a user and the user's neighbors leads to a decline in\nengagement. Our analysis has practical implications for improving link\nrecommender systems.\n",
"title": "Beautiful and damned. Combined effect of content quality and social ties on user engagement"
}
| null | null |
[
"Computer Science"
] | null | true | null |
15604
| null |
Validated
| null | null |
null |
{
"abstract": " The detection of software vulnerabilities (or vulnerabilities for short) is\nan important problem that has yet to be tackled, as manifested by many\nvulnerabilities reported on a daily basis. This calls for machine learning\nmethods to automate vulnerability detection. Deep learning is attractive for\nthis purpose because it does not require human experts to manually define\nfeatures. Despite the tremendous success of deep learning in other domains, its\napplicability to vulnerability detection is not systematically understood. In\norder to fill this void, we propose the first systematic framework for using\ndeep learning to detect vulnerabilities. The framework, dubbed Syntax-based,\nSemantics-based, and Vector Representations (SySeVR), focuses on obtaining\nprogram representations that can accommodate syntax and semantic information\npertinent to vulnerabilities. Our experiments with 4 software products\ndemonstrate the usefulness of the framework: we detect 15 vulnerabilities that\nare not reported in the National Vulnerability Database. Among these 15\nvulnerabilities, 7 are unknown and have been reported to the vendors, and the\nother 8 have been \"silently\" patched by the vendors when releasing newer\nversions of the products.\n",
"title": "SySeVR: A Framework for Using Deep Learning to Detect Software Vulnerabilities"
}
| null | null | null | null | true | null |
15605
| null |
Default
| null | null |
null |
{
"abstract": " We investigate the problem of learning discrete, undirected graphical models\nin a differentially private way. We show that the approach of releasing noisy\nsufficient statistics using the Laplace mechanism achieves a good trade-off\nbetween privacy, utility, and practicality. A naive learning algorithm that\nuses the noisy sufficient statistics \"as is\" outperforms general-purpose\ndifferentially private learning algorithms. However, it has three limitations:\nit ignores knowledge about the data generating process, rests on uncertain\ntheoretical foundations, and exhibits certain pathologies. We develop a more\nprincipled approach that applies the formalism of collective graphical models\nto perform inference over the true sufficient statistics within an\nexpectation-maximization framework. We show that this learns better models than\ncompeting approaches on both synthetic data and on real human mobility data\nused as a case study.\n",
"title": "Differentially Private Learning of Undirected Graphical Models using Collective Graphical Models"
}
| null | null | null | null | true | null |
15606
| null |
Default
| null | null |
null |
{
"abstract": " We consider orthogonal decompositions of invariant subspaces of Hardy spaces,\nthese relate to the Blaschke based phase unwinding decompositions. We prove\nconvergence in Lp. In particular we build an explicit multiscale wavelet basis.\nWe also obtain an explicit unwindinig decomposition for the singular inner\nfunction, exp 2i\\pi/x.\n",
"title": "Phase unwinding, or invariant subspace decompositions of Hardy spaces"
}
| null | null | null | null | true | null |
15607
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we propose a novel object proposal generation scheme by\nformulating a graph-based salient edge classification framework that utilizes\nthe edge context. In the proposed method, we construct a Bayesian probabilistic\nedge map to assign a saliency value to the edgelets by exploiting low level\nedge features. A Conditional Random Field is then learned to effectively\ncombine these features for edge classification with object/non-object label. We\npropose an objectness score for the generated windows by analyzing the salient\nedge density inside the bounding box. Extensive experiments on PASCAL VOC 2007\ndataset demonstrate that the proposed method gives competitive performance\nagainst 10 popular generic object detection techniques while using fewer number\nof proposals.\n",
"title": "SalProp: Salient object proposals via aggregated edge cues"
}
| null | null |
[
"Computer Science"
] | null | true | null |
15608
| null |
Validated
| null | null |
null |
{
"abstract": " We study the effects of quantum fluctuations on the dynamical generation of a\ngap and on the evolution of the spin-wave spectra of a frustrated magnet on a\ntriangular lattice with bond-dependent Ising couplings, analog of the Kitaev\nhoneycomb model. The quantum fluctuations lift the subextensive degeneracy of\nthe classical ground-state manifold by a quantum order-by-disorder mechanism.\nNearest-neighbor chains remain decoupled and the surviving discrete degeneracy\nof the ground state is protected by a hidden model symmetry. We show how the\nfour-spin interaction, emergent from the fluctuations, generates a spin gap\nshifting the nodal lines of the linear spin-wave spectrum to finite energies.\n",
"title": "Quantum gap and spin-wave excitations in the Kitaev model on a triangular lattice"
}
| null | null | null | null | true | null |
15609
| null |
Default
| null | null |
null |
{
"abstract": " Independent component analysis (ICA) is a cornerstone of modern data\nanalysis. Its goal is to recover a latent random vector S with independent\ncomponents from samples of X=AS where A is an unknown mixing matrix.\nCritically, all existing methods for ICA rely on and exploit strongly the\nassumption that S is not Gaussian as otherwise A becomes unidentifiable. In\nthis paper, we show that in fact one can handle the case of Gaussian components\nby imposing structure on the matrix A. Specifically, we assume that A is sparse\nand generic in the sense that it is generated from a sparse Bernoulli-Gaussian\nensemble. Under this condition, we give an efficient algorithm to recover the\ncolumns of A given only the covariance matrix of X as input even when S has\nseveral Gaussian components.\n",
"title": "Sparse Gaussian ICA"
}
| null | null | null | null | true | null |
15610
| null |
Default
| null | null |
null |
{
"abstract": " As a general and thus popular model for autonomous systems, partially\nobservable Markov decision process (POMDP) can capture uncertainties from\ndifferent sources like sensing noises, actuation errors, and uncertain\nenvironments. However, its comprehensiveness makes the planning and control in\nPOMDP difficult. Traditional POMDP planning problems target to find the optimal\npolicy to maximize the expectation of accumulated rewards. But for safety\ncritical applications, guarantees of system performance described by formal\nspecifications are desired, which motivates us to consider formal methods to\nsynthesize supervisor for POMDP. With system specifications given by\nProbabilistic Computation Tree Logic (PCTL), we propose a supervisory control\nframework with a type of deterministic finite automata (DFA), za-DFA, as the\ncontroller form. While the existing work mainly relies on optimization\ntechniques to learn fixed-size finite state controllers (FSCs), we develop an\n$L^*$ learning based algorithm to determine both space and transitions of\nza-DFA. Membership queries and different oracles for conjectures are defined.\nThe learning algorithm is sound and complete. An example is given in detailed\nsteps to illustrate the supervisor synthesis algorithm.\n",
"title": "Supervisor Synthesis of POMDP based on Automata Learning"
}
| null | null | null | null | true | null |
15611
| null |
Default
| null | null |
null |
{
"abstract": " In 2015, Barber and Candes introduced a new variable selection procedure\ncalled the knockoff filter to control the false discovery rate (FDR) and prove\nthat this method achieves exact FDR control. Inspired by the work of Barber and\nCandes (2015), we propose and analyze a pseudo-knockoff filter that inherits\nsome advantages of the original knockoff filter and has more flexibility in\nconstructing its knockoff matrix. Although we have not been able to obtain\nexact FDR control of the pseudo knockoff filter, we show that it satisfies an\nexpectation inequality that offers some insight into FDR control. Moreover, we\nprovide some partial analysis of the pseudo knockoff filter for the half Lasso\nand the least squares statistics. Our analysis indicates that the inverse of\nthe covariance matrix of the feature matrix plays an important role in\ndesigning and analyzing the pseudo knockoff filter. Our preliminary numerical\nexperiments show that the pseudo knockoff filter with the half Lasso statistic\nhas FDR control. Moreover, our numerical experiments show that the\npseudo-knockoff filter could offer more power than the original knockoff filter\nwith the OMP or Lasso Path statistic when the features are correlated and\nnon-sparse.\n",
"title": "A Pseudo Knockoff Filter for Correlated Features"
}
| null | null | null | null | true | null |
15612
| null |
Default
| null | null |
null |
{
"abstract": " This paper is a sequel to [He11] and [GH17]. In [He11] a notion of marking of\nisolated hypersurface singularities was defined, and a moduli space\n$M_\\mu^{mar}$ for marked singularities in one $\\mu$-homotopy class of isolated\nhypersurface singularities was established. It is an analogue of a\nTeichmüller space. It comes together with a $\\mu$-constant monodromy group\n$G^{mar}\\subset G_{\\mathbb{Z}}$. Here $G_{\\mathbb{Z}}$ is the group of\nautomorphisms of a Milnor lattice which respect the Seifert form. It was\nconjectured that $M_\\mu^{mar}$ is connected. This is equivalent to $G^{mar}=\nG_{\\mathbb{Z}}$. Also Torelli type conjectures were formulated. In [He11] and\n[GH17] $M_\\mu^{mar}, G_{\\mathbb{Z}}$ and $G^{mar}$ were determined and all\nconjectures were proved for the simple, the unimodal and the exceptional\nbimodal singularities. In this paper the quadrangle singularities and the\nbimodal series are treated. The Torelli type conjectures are true. But the\nconjecture $G^{mar}= G_{\\mathbb{Z}}$ and $M_\\mu^{mar}$ connected does not hold\nfor certain subseries of the bimodal series.\n",
"title": "$μ$-constant monodromy groups and Torelli results for the quadrangle singularities and the bimodal series"
}
| null | null |
[
"Mathematics"
] | null | true | null |
15613
| null |
Validated
| null | null |
null |
{
"abstract": " We propose SoaAlloc, a dynamic object allocator for Single-Method\nMultiple-Objects applications in CUDA. SoaAlloc is the first allocator for GPUs\nthat (a) arranges allocations in a SIMD-friendly Structure of Arrays (SOA) data\nlayout, (b) provides a do-all operation for maximizing the benefit of SOA, and\n(c) is on par with state-of-the-art memory allocators for raw (de)allocation\ntime. Our benchmarks show that the SOA layout leads to significantly better\nmemory bandwidth utilization, resulting in a 2x speedup of application code.\n",
"title": "SoaAlloc: Accelerating Single-Method Multiple-Objects Applications on GPUs"
}
| null | null |
[
"Computer Science"
] | null | true | null |
15614
| null |
Validated
| null | null |
null |
{
"abstract": " The radiological characterization of contaminated elements (walls, grounds,\nobjects) from nuclear facilities often suffers from a too small number of\nmeasurements. In order to determine risk prediction bounds on the level of\ncontamination, some classic statistical methods may then reveal unsuited as\nthey rely upon strong assumptions (e.g. that the underlying distribution is\nGaussian) which cannot be checked. Considering that a set of measurements or\ntheir average value arise from a Gaussian distribution can sometimes lead to\nerroneous conclusion, possibly underconservative. This paper presents several\nalternative statistical approaches which are based on much weaker hypotheses\nthan Gaussianity. They result from general probabilistic inequalities and\norder-statistics based formula. Given a data sample, these inequalities make it\npossible to derive prediction intervals for a random variable, which can be\ndirectly interpreted as probabilistic risk bounds. For the sake of validation,\nthey are first applied to synthetic data samples generated from several known\ntheoretical distributions. In a second time, the proposed methods are applied\nto two data sets obtained from real radiological contamination measurements.\n",
"title": "Probabilistic risk bounds for the characterization of radiological contamination"
}
| null | null | null | null | true | null |
15615
| null |
Default
| null | null |
null |
{
"abstract": " In this note we derive the backward (automatic) differentiation (adjoint\n[automatic] differentiation) for an algorithm containing a conditional\nexpectation operator. As an example we consider the backward algorithm as it is\nused in Bermudan product valuation, but the method is applicable in full\ngenerality.\nThe method relies on three simple properties: 1) a forward or backward\n(automatic) differentiation of an algorithm containing a conditional\nexpectation operator results in a linear combination of the conditional\nexpectation operators; 2) the differential of an expectation is the expectation\nof the differential $\\frac{d}{dx} E(Y) = E(\\frac{d}{dx}Y)$; 3) if we are only\ninterested in the expectation of the final result (as we are in all valuation\nproblems), we may use $E(A \\cdot E(B\\vert\\mathcal{F})) = E(E(A\\vert\\mathcal{F})\n\\cdot B)$, i.e., instead of applying the (conditional) expectation operator to\na function of the underlying random variable (continuation values), it may be\napplied to the adjoint differential. \\end{enumerate}\nThe methodology not only allows for a very clean and simple implementation,\nbut also offers the ability to use different conditional expectation estimators\nin the valuation and the differentiation.\n",
"title": "Automatic Backward Differentiation for American Monte-Carlo Algorithms (Conditional Expectation)"
}
| null | null | null | null | true | null |
15616
| null |
Default
| null | null |
null |
{
"abstract": " Fundamental atomic parameters, such as oscillator strengths, play a key role\nin modelling and understanding the chemical composition of stars in the\nuniverse. Despite the significant work underway to produce these parameters for\nmany astrophysically important ions, uncertainties in these parameters remain\nlarge and can propagate throughout the entire field of astronomy. The Belgian\nrepository of fundamental atomic data and stellar spectra (BRASS) aims to\nprovide the largest systematic and homogeneous quality assessment of atomic\ndata to date in terms of wavelength, atomic and stellar parameter coverage. To\nprepare for it, we first compiled multiple literature occurrences of many\nindividual atomic transitions, from several atomic databases of astrophysical\ninterest, and assessed their agreement. Several atomic repositories were\nsearched and their data retrieved and formatted in a consistent manner. Data\nentries from all repositories were cross-matched against our initial BRASS\natomic line list to find multiple occurrences of the same transition. Where\npossible we used a non-parametric cross-match depending only on electronic\nconfigurations and total angular momentum values. We also checked for duplicate\nentries of the same physical transition, within each retrieved repository,\nusing the non-parametric cross-match. We report the cross-matched transitions\nfor each repository and compare their fundamental atomic parameters. We find\ndifferences in log(gf) values of up to 2 dex or more. We also find and report\nthat ~2% of our line list and Vienna Atomic Line Database retrievals are\ncomposed of duplicate transitions. Finally we provide a number of examples of\natomic spectral lines with different log(gf) values, and discuss the impact of\nthese uncertain log(gf) values on quantitative spectroscopy. All cross-matched\natomic data and duplicate transitions are available to download at\nbrass.sdf.org.\n",
"title": "The Belgian repository of fundamental atomic data and stellar spectra (BRASS). I. Cross-matching atomic databases of astrophysical interest"
}
| null | null |
[
"Physics"
] | null | true | null |
15617
| null |
Validated
| null | null |
null |
{
"abstract": " Hamiltonian Truncation (a.k.a. Truncated Spectrum Approach) is an efficient\nnumerical technique to solve strongly coupled QFTs in d=2 spacetime dimensions.\nFurther theoretical developments are needed to increase its accuracy and the\nrange of applicability. With this goal in mind, here we present a new variant\nof Hamiltonian Truncation which exhibits smaller dependence on the UV cutoff\nthan other existing implementations, and yields more accurate spectra. The key\nidea for achieving this consists in integrating out exactly a certain class of\nhigh energy states, which corresponds to performing renormalization at the\ncubic order in the interaction strength. We test the new method on the strongly\ncoupled two-dimensional quartic scalar theory. Our work will also be useful for\nthe future goal of extending Hamiltonian Truncation to higher dimensions d >=\n3.\n",
"title": "High-Precision Calculations in Strongly Coupled Quantum Field Theory with Next-to-Leading-Order Renormalized Hamiltonian Truncation"
}
| null | null |
[
"Physics"
] | null | true | null |
15618
| null |
Validated
| null | null |
null |
{
"abstract": " A critical and challenging problem in reinforcement learning is how to learn\nthe state-action value function from the experience replay buffer and\nsimultaneously keep sample efficiency and faster convergence to a high quality\nsolution. In prior works, transitions are uniformly sampled at random from the\nreplay buffer or sampled based on their priority measured by\ntemporal-difference (TD) error. However, these approaches do not fully take\ninto consideration the intrinsic characteristics of transition distribution in\nthe state space and could result in redundant and unnecessary TD updates,\nslowing down the convergence of the learning procedure. To overcome this\nproblem, we propose a novel state distribution-aware sampling method to balance\nthe replay times for transitions with skew distribution, which takes into\naccount both the occurrence frequencies of transitions and the uncertainty of\nstate-action values. Consequently, our approach could reduce the unnecessary TD\nupdates and increase the TD updates for state-action value with more\nuncertainty, making the experience replay more effective and efficient.\nExtensive experiments are conducted on both classic control tasks and Atari\n2600 games based on OpenAI gym platform and the experimental results\ndemonstrate the effectiveness of our approach in comparison with the standard\nDQN approach.\n",
"title": "State Distribution-aware Sampling for Deep Q-learning"
}
| null | null | null | null | true | null |
15619
| null |
Default
| null | null |
null |
{
"abstract": " High penetration of renewable energy source makes microgrid (MGs) be\nenvironment friendly. However, the stochastic input from renewable energy\nresource brings difficulty in balancing the energy supply and demand.\nPurchasing extra energy from macrogrid to deal with energy shortage will\nincrease MG energy cost. To mitigate intermittent nature of renewable energy,\nenergy trading and energy storage which can exploit diversity of renewable\nenergy generation across space and time are efficient and cost-effective\nmethods. But current energy storage control action will impact the future\ncontrol action which brings challenge to energy management. In addition, due to\nMG participating energy trading as prosumer, it calls for an efficient trading\nmechanism. Therefore, this paper focuses on the problem of MG energy management\nand trading. Energy trading problem is formulated as a stochastic optimization\none with both individual profit and social welfare maximization. Firstly a\nLyapunov optimization based algorithm is developed to solve the stochastic\nproblem. Secondly the double-auction based mechanism is provided to attract MG\ntruthful bidding for buying and selling energy. Through theoretical analysis,\nwe demonstrate that individual MG can achieve a time average energy cost close\nto offline optimum with tradeoff between storage capacity and energy trading\ncost. Meanwhile the social welfare is also asymptotically maximized under\ndouble auction. Simulation results based on real world data show the\neffectiveness of our algorithm.\n",
"title": "Energy Trading between microgrids Individual Cost Minimization and Social Welfare Maximization"
}
| null | null | null | null | true | null |
15620
| null |
Default
| null | null |
null |
{
"abstract": " We summarize our recent findings, where we proposed a framework for learning\na Kolmogorov model, for a collection of binary random variables. More\nspecifically, we derive conditions that link outcomes of specific random\nvariables, and extract valuable relations from the data. We also propose an\nalgorithm for computing the model and show its first-order optimality, despite\nthe combinatorial nature of the learning problem. We apply the proposed\nalgorithm to recommendation systems, although it is applicable to other\nscenarios. We believe that the work is a significant step toward interpretable\nmachine learning.\n",
"title": "Learning Kolmogorov Models for Binary Random Variables"
}
| null | null | null | null | true | null |
15621
| null |
Default
| null | null |
null |
{
"abstract": " Intensive studies for more than three decades have elucidated multiple\nsuperconducting phases and odd-parity Cooper pairs in a heavy fermion\nsuperconductor UPt$_3$. We identify a time-reversal invariant superconducting\nphase of UPt$_3$ as a recently proposed topological nonsymmorphic\nsuperconductivity. Combining the band structure of UPt$_3$, order parameter of\n$E_{\\rm 2u}$ representation allowed by $P6_3/mmc$ space group symmetry, and\ntopological classification by $K$-theory, we demonstrate the nontrivial\n$Z_2$-invariant of three-dimensional DIII class enriched by glide symmetry.\nCorrespondingly, double Majorana cone surface states appear at the surface\nBrillouin zone boundary. Furthermore, we show a variety of surface states and\nclarify the topological protection by crystal symmetry. Majorana arcs\ncorresponding to tunable Weyl points appear in the time-reversal symmetry\nbroken B-phase. Majorana cone protected by mirror Chern number and Majorana\nflat band by glide-winding number are also revealed.\n",
"title": "Möbius topological superconductivity in UPt$_3$"
}
| null | null | null | null | true | null |
15622
| null |
Default
| null | null |
null |
{
"abstract": " The Vector AutoRegressive Moving Average (VARMA) model is fundamental to the\nstudy of multivariate time series. However, estimation becomes challenging in\neven relatively low-dimensional VARMA models. With growing interest in the\nsimultaneous modeling of large numbers of marginal time series, many authors\nhave abandoned the VARMA model in favor of the Vector AutoRegressive (VAR)\nmodel, which is seen as a simpler alternative, both in theory and practice, in\nthis high-dimensional context. However, even very simple VARMA models can be\nvery complicated to represent using only VAR modeling. In this paper, we\ndevelop a new approach to VARMA identification and propose a two-phase method\nfor estimation. Our identification and estimation strategies are linked in\ntheir use of sparsity-inducing convex regularizers, which favor VARMA models\nthat have only a small number of nonzero parameters. We establish sufficient\nconditions for consistency of sparse infinite-order VAR estimates in high\ndimensions, a key ingredient for our two-phase sparse VARMA estimation\nstrategy. The proposed framework has good estimation and forecast accuracy\nunder numerous simulation settings. We illustrate the forecast performance of\nthe sparse VARMA models for several application domains, including\nmacro-economic forecasting, demand forecasting, and volatility forecasting. The\nproposed sparse VARMA estimator gives parsimonious forecast models that lead to\nimportant gains in relative forecast accuracy.\n",
"title": "Sparse Identification and Estimation of High-Dimensional Vector AutoRegressive Moving Averages"
}
| null | null | null | null | true | null |
15623
| null |
Default
| null | null |
null |
{
"abstract": " By investigating information flow between a general parity-time (PT)\n-symmetric non-Hermitian system and an environment, we find that the complete\ninformation retrieval from the environment can be achieved in the PT-unbroken\nphase, whereas no information can be retrieved in the PT-broken phase. The\nPT-transition point thus marks the reversible-irreversible criticality of\ninformation flow, around which many physical quantities such as the recurrence\ntime and the distinguishability between quantum states exhibit power-law\nbehavior. Moreover, by embedding a PT-symmetric system into a larger Hilbert\nspace so that the entire system obeys unitary dynamics, we reveal that behind\nthe information retrieval lies a hidden entangled partner protected by PT\nsymmetry. Possible experimental situations are also discussed.\n",
"title": "Information Retrieval and Criticality in Parity-Time-Symmetric Systems"
}
| null | null | null | null | true | null |
15624
| null |
Default
| null | null |
null |
{
"abstract": " Most popular strategies to capture subjective judgments from humans involve\nthe construction of a unidimensional relative measurement scale, representing\norder preferences or judgments about a set of objects or conditions. This\ninformation is generally captured by means of direct scoring, either in the\nform of a Likert or cardinal scale, or by comparative judgments in pairs or\nsets. In this sense, the use of pairwise comparisons is becoming increasingly\npopular because of the simplicity of this experimental procedure. However, this\nstrategy requires non-trivial data analysis to aggregate the comparison ranks\ninto a quality scale and analyse the results, in order to take full advantage\nof the collected data. This paper explains the process of translating pairwise\ncomparison data into a measurement scale, discusses the benefits and\nlimitations of such scaling methods and introduces a publicly available\nsoftware in Matlab. We improve on existing scaling methods by introducing\noutlier analysis, providing methods for computing confidence intervals and\nstatistical testing and introducing a prior, which reduces estimation error\nwhen the number of observers is low. Most of our examples focus on image\nquality assessment.\n",
"title": "A practical guide and software for analysing pairwise comparison experiments"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
15625
| null |
Validated
| null | null |
null |
{
"abstract": " The world is connected through the Internet. As the abundance of Internet\nusers connected into the Web and the popularity of cloud computing research,\nthe need of Artificial Intelligence (AI) is demanding. In this research,\nGenetic Algorithm (GA) as AI optimization method through natural selection and\ngenetic evolution is utilized. There are many applications of GA such as web\nmining, load balancing, routing, and scheduling or web service selection.\nHence, it is a challenging task to discover whether the code mainly server side\nand web based language technology affects the performance of GA. Travelling\nSalesperson Problem (TSP) as Non Polynomial-hard (NP-hard) problem is provided\nto be a problem domain to be solved by GA. While many scientists prefer Python\nin GA implementation, another popular high-level interpreter programming\nlanguage such as PHP (PHP Hypertext Preprocessor) and Ruby were benchmarked.\nLine of codes, file sizes, and performances based on GA implementation and\nruntime were found varies among these programming languages. Based on the\nresult, the use of Ruby in GA implementation is recommended.\n",
"title": "Web-Based Implementation of Travelling Salesperson Problem Using Genetic Algorithm"
}
| null | null | null | null | true | null |
15626
| null |
Default
| null | null |
null |
{
"abstract": " Random network models play a prominent role in modeling, analyzing and\nunderstanding complex phenomena on real-life networks. However, a key property\nof networks is often neglected: many real-world networks exhibit spatial\nstructure, the tendency of a node to select neighbors with a probability\ndepending on physical distance. Here, we introduce a class of random spatial\nnetworks (RSNs) which generalizes many existing random network models but adds\nspatial structure. In these networks, nodes are placed randomly in space and\njoined in edges with a probability depending on their distance and their\nindividual expected degrees, in a manner that crucially remains analytically\ntractable. We use this network class to propose a new generalization of\nsmall-world networks, where the average shortest path lengths in the graph are\nsmall, as in classical Watts-Strogatz small-world networks, but with close\nspatial proximity of nodes that are neighbors in the network playing the role\nof large clustering. Small-world effects are demonstrated on these spatial\nsmall-world networks without clustering. We are able to derive partial\nintegro-differential equations governing susceptible-infectious-recovered\ndisease spreading through an RSN, and we demonstrate the existence of traveling\nwave solutions. If the distance kernel governing edge placement decays slower\nthan exponential, the population-scale dynamics are dominated by long-range\nhops followed by local spread of traveling waves. This provides a theoretical\nmodeling framework for recent observations of how epidemics like Ebola evolve\nin modern connected societies, with long-range connections seeding new focal\npoints from which the epidemic locally spreads in a wavelike manner.\n",
"title": "Random Spatial Networks: Small Worlds without Clustering, Traveling Waves, and Hop-and-Spread Disease Dynamics"
}
| null | null | null | null | true | null |
15627
| null |
Default
| null | null |
null |
{
"abstract": " Large-scale dipolar surface magnetic fields have been detected in a fraction\nof OB stars, however only few stellar evolution models of massive stars have\nconsidered the impact of these fossil fields. We are performing 1D\nhydrodynamical model calculations taking into account evolutionary consequences\nof the magnetospheric-wind interactions in a simplified parametric way. Two\neffects are considered: i) the global mass-loss rates are reduced due to\nmass-loss quenching, and ii) the surface angular momentum loss is enhanced due\nto magnetic braking. As a result of the magnetic mass-loss quenching, the mass\nof magnetic massive stars remains close to their initial masses. Thus magnetic\nmassive stars - even at Galactic metallicity - have the potential to be\nprogenitors of `heavy' stellar mass black holes. Similarly, at Galactic\nmetallicity, the formation of pair instability supernovae is plausible with a\nmagnetic progenitor.\n",
"title": "The evolution of magnetic hot massive stars: Implementation of the quantitative influence of surface magnetic fields in modern models of stellar evolution"
}
| null | null | null | null | true | null |
15628
| null |
Default
| null | null |
null |
{
"abstract": " The focusing NLS equation is the simplest universal model describing the\nmodulation instability (MI) of quasi monochromatic waves in weakly nonlinear\nmedia, considered the main physical mechanism for the appearance of rogue\n(anomalous) waves (RWs) in Nature. In this paper we study, using the finite gap\nmethod, the NLS Cauchy problem for periodic initial perturbations of the\nunstable background solution of NLS exciting just one of the unstable modes. We\ndistinguish two cases. In the case in which only the corresponding unstable gap\nis theoretically open, the solution describes an exact deterministic alternate\nrecurrence of linear and nonlinear stages of MI, and the nonlinear RW stages\nare described by the 1-breather Akhmediev solution, whose parameters, different\nat each RW appearance, are always given in terms of the initial data through\nelementary functions. If the number of unstable modes is >1, this uniform in t\ndynamics is sensibly affected by perturbations due to numerics and/or real\nexperiments, provoking O(1) corrections to the result. In the second case in\nwhich more than one unstable gap is open, a detailed investigation of all these\ngaps is necessary to get a uniform in $t$ dynamics, and this study is postponed\nto a subsequent paper. It is however possible to obtain the elementary\ndescription of the first nonlinear stage of MI, given again by the Akhmediev\n1-breather solution, and how perturbations due to numerics and/or real\nexperiments can affect this result.\n",
"title": "The finite gap method and the analytic description of the exact rogue wave recurrence in the periodic NLS Cauchy problem. 1"
}
| null | null | null | null | true | null |
15629
| null |
Default
| null | null |
null |
{
"abstract": " We present the first discoveries from a survey of $z\\gtrsim6$ quasars using\nimaging data from the DECam Legacy Survey (DECaLS) in the optical, the UKIRT\nDeep Infrared Sky Survey (UKIDSS) and a preliminary version of the UKIRT\nHemisphere Survey (UHS) in the near-IR, and ALLWISE in the mid-IR. DECaLS will\nimage 9000 deg$^2$ of sky down to $z_{\\rm AB}\\sim23.0$, and UKIDSS and UHS,\nwhich will map the northern sky at $0<DEC<+60^{\\circ}$, reaching $J_{\\rm\nVEGA}\\sim19.6$ (5-$\\sigma$). The combination of these datasets allows us to\ndiscover quasars at redshift $z\\gtrsim7$ and to conduct a complete census of\nthe faint quasar population at $z\\gtrsim6$. In this paper, we report on the\nselection method of our search, and on the initial discoveries of two new,\nfaint $z\\gtrsim6$ quasars and one new $z=6.63$ quasar in our pilot\nspectroscopic observations. The two new $z\\sim6$ quasars are at $z=6.07$ and\n$z=6.17$ with absolute magnitudes at rest-frame wavelength 1450 \\AA\\ being\n$M_{1450}=-25.83$ and $M_{1450}=-25.76$, respectively. These discoveries\nsuggest that we can find quasars close to or fainter than the break magnitude\nof the Quasar Luminosity Function (QLF) at $z\\gtrsim6$. The new $z=6.63$ quasar\nhas an absolute magnitude of $M_{1450}=-25.95$. This demonstrates the potential\nof using the combined DECaLS and UKIDSS/UHS datasets to find $z\\gtrsim7$\nquasars. Extrapolating from previous QLF measurements, we predict that these\ncombined datasets will yield $\\sim200$ $z\\sim6$ quasars to $z_{\\rm AB} < 21.5$,\n$\\sim1{,}000$ $z\\sim6$ quasars to $z_{\\rm AB}<23$, and $\\sim 30$ quasars at\n$z>6.5$ to $J_{\\rm VEGA}<19.5$.\n",
"title": "First Discoveries of z>6 Quasars with the DECam Legacy Survey and UKIRT Hemisphere Survey"
}
| null | null | null | null | true | null |
15630
| null |
Default
| null | null |
null |
{
"abstract": " We prove that if a solution of the time-dependent Schr{ö}dinger equation on\nan homogeneous tree with bounded potential decays fast at two distinct times\nthen the solution is trivial. For the free Schr{ö}dinger operator, we use the\nspectral theory of the Laplacian and complex analysis and obtain a\ncharacterization of the initial conditions that lead to a sharp decay at any\ntime. We then use the recent spectral decomposition of the Schr{ö}dinger\noperator with compactly supported potential due to Colin de Verdi{è}rre and\nTurc to extend our results in the presence of such potentials. Finally, we use\nreal variable methods first introduced by Escauriaza, Kenig, Ponce and Vega to\nestablish a general sharp result in the case of bounded potentials.\n",
"title": "On unique continuation for solutions of the Schr{ö}dinger equation on trees"
}
| null | null | null | null | true | null |
15631
| null |
Default
| null | null |
null |
{
"abstract": " One of the most puzzling features of high-temperature cuprate superconductors\nis the pseudogap state, which appears above the temperature at which\nsuperconductivity is destroyed. There remain fundamental questions regarding\nits nature and its relation to superconductivity. But to address these\nquestions, we must first determine whether the pseudogap and superconducting\nstates share a common property: particle-hole symmetry. We introduce a new\ntechnique to test particle-hole symmetry by using laser pulses to manipulate\nand measure the chemical potential on picosecond time scales. The results\nstrongly suggest that the asymmetry in the density of states is inverted in the\npseudogap state, implying a particle-hole asymmetric gap. Independent of\ninterpretation, these results can test theoretical predictions of the density\nof states in cuprates.\n",
"title": "Particle-hole Asymmetry in the Cuprate Pseudogap Measured with Time-Resolved Spectroscopy"
}
| null | null | null | null | true | null |
15632
| null |
Default
| null | null |
null |
{
"abstract": " Thermal noise is expected to be one of the noise sources limiting the\nastrophysical reach of Advanced LIGO (once commissioning is complete) and\nthird-generation detectors. Adopting crystalline materials for thin, reflecting\nmirror coatings, rather than the amorphous coatings used in current-generation\ndetectors, could potentially reduce thermal noise. Understanding and reducing\nthermal noise requires accurate theoretical models, but modeling thermal noise\nanalytically is especially challenging with crystalline materials. Thermal\nnoise models typically rely on the fluctuation-dissipation theorem, which\nrelates the power spectral density of the thermal noise to an auxiliary elastic\nproblem. In this paper, we present results from a new, open-source tool that\nnumerically solves the auxiliary elastic problem to compute the Brownian\nthermal noise for both amorphous and crystalline coatings. We employ\nopen-source frameworks to solve the auxiliary elastic problem using a\nfinite-element method, adaptive mesh refinement, and parallel processing that\nenables us to use high resolutions capable of resolving the thin reflective\ncoating. We compare with approximate analytic solutions for amorphous\nmaterials, and we verify that our solutions scale as expected. Finally, we\nmodel the crystalline coating thermal noise in an experiment reported by Cole\nand collaborators (2013), comparing our results to a simpler numerical\ncalculation that treats the coating as an \"effectively amorphous\" material. We\nfind that treating the coating as a cubic crystal instead of as an effectively\namorphous material increases the thermal noise by about 3%. Our results are a\nstep toward better understanding and reducing thermal noise to increase the\nreach of future gravitational-wave detectors. (Abstract abbreviated.)\n",
"title": "Numerically modeling Brownian thermal noise in amorphous and crystalline thin coatings"
}
| null | null | null | null | true | null |
15633
| null |
Default
| null | null |
null |
{
"abstract": " In an effort to understand the meaning of the intermediate representations\ncaptured by deep networks, recent papers have tried to associate specific\nsemantic concepts to individual neural network filter responses, where\ninteresting correlations are often found, largely by focusing on extremal\nfilter responses. In this paper, we show that this approach can favor\neasy-to-interpret cases that are not necessarily representative of the average\nbehavior of a representation.\nA more realistic but harder-to-study hypothesis is that semantic\nrepresentations are distributed, and thus filters must be studied in\nconjunction. In order to investigate this idea while enabling systematic\nvisualization and quantification of multiple filter responses, we introduce the\nNet2Vec framework, in which semantic concepts are mapped to vectorial\nembeddings based on corresponding filter responses. By studying such\nembeddings, we are able to show that 1., in most cases, multiple filters are\nrequired to code for a concept, that 2., often filters are not concept specific\nand help encode multiple concepts, and that 3., compared to single filter\nactivations, filter embeddings are able to better characterize the meaning of a\nrepresentation and its relationship to other concepts.\n",
"title": "Net2Vec: Quantifying and Explaining how Concepts are Encoded by Filters in Deep Neural Networks"
}
| null | null | null | null | true | null |
15634
| null |
Default
| null | null |
null |
{
"abstract": " We investigate how star formation efficiency can be significantly decreased\nby the removal of a molecular cloud's envelope by feedback from an external\nsource. Feedback from star formation has difficulties halting the process in\ndense gas but can easily remove the less dense and warmer envelopes where star\nformation does not occur. However, the envelopes can play an important role\nkeeping their host clouds bound by deepening the gravitational potential and\nproviding a constraining pressure boundary. We use numerical simulations to\nshow that removal of the cloud envelopes results in all cases in a fall in the\nstar formation efficiency (SFE). At 1.38 free-fall times our 4 pc cloud\nsimulation experienced a drop in the SFE from 16 to six percent, while our 5 pc\ncloud fell from 27 to 16 per cent. At the same time, our 3 pc cloud (the least\nbound) fell from an SFE of 5.67 per cent to zero when the envelope was lost.\nThe star formation efficiency per free-fall time varied from zero to $\\approx$\n0.25 according to $\\alpha$, defined to be the ratio of the kinetic plus thermal\nto gravitational energy, and irrespective of the absolute star forming mass\navailable. Furthermore the fall in SFE associated with the loss of the envelope\nis found to even occur at later times. We conclude that the SFE will always\nfall should a star forming cloud lose its envelope due to stellar feedback,\nwith less bound clouds suffering the greatest decrease.\n",
"title": "Can the removal of molecular cloud envelopes by external feedback affect the efficiency of star formation?"
}
| null | null | null | null | true | null |
15635
| null |
Default
| null | null |
null |
{
"abstract": " In hybrid digital-analog (HDA) systems, resource allocation has been utilized\nto achieve desired distortion performance. However, existing studies on this\nissue assume error-free digital transmission, which is not valid for fading\nchannels. With time-varying channel fading, the exact channel state information\nis not available at the transmitter. Thus, random outage and resulting digital\ndistortion cannot be ignored. Moreover, rate allocation should be considered in\nresource allocation, since it not only determines the amount of information for\ndigital transmission and that for analog transmission, but also affects the\noutage probability. Based on above observations, in this paper, we attempt to\nperform joint rate and resource allocation strategies to optimize system\ndistortion in HDA systems over fading channels. Consider a bandwidth expansion\nscenario where a memoryless Gaussian source is transmitted in an HDA system\nwith the entropy-constrained scalar quantizer (ECSQ). Firstly, we formulate the\njoint allocation problem as an expected system distortion minimization problem\nwhere both analog and digital distortion are considered. Then, in the limit of\nlow outage probability, we decompose the problem into two coupled sub-problems\nbased on the block coordinate descent method, and propose an iterative gradient\nalgorithm to approach the optimal solution. Furthermore, we extend our work to\nthe multivariate Gaussian source scenario where a two-stage fast algorithm\nintegrating rounding and greedy strategies is proposed to optimize the joint\nrate and resource allocation problem. Finally, simulation results demonstrate\nthat the proposed algorithms can achieve up to 2.3dB gains in terms of\nsignal-to-distortion ratio over existing schemes under the single Gaussian\nsource scenario, and up to 3.5dB gains under the multivariate Gaussian source\nscenario.\n",
"title": "Joint Rate and Resource Allocation in Hybrid Digital-Analog Transmission over Fading Channels"
}
| null | null | null | null | true | null |
15636
| null |
Default
| null | null |
null |
{
"abstract": " We present an experimental study on the non-equilibrium tunnel dynamics of\ntwo coupled one-dimensional Bose-Einstein quasi-condensates deep in the\nJosephson regime. Josephson oscillations are initiated by splitting a single\none-dimensional condensate and imprinting a relative phase between the\nsuperfluids. Regardless of the initial state and experimental parameters, the\ndynamics of the relative phase and atom number imbalance shows a relaxation to\na phase-locked steady state. The latter is characterized by a high phase\ncoherence and reduced fluctuations with respect to the initial state. We\npropose an empirical model based on the analogy with the anharmonic oscillator\nto describe the effect of various experimental parameters. A microscopic theory\ncompatible with our observations is still missing.\n",
"title": "Relaxation to a Phase-locked Equilibrium State in a One-dimensional Bosonic Josephson Junction"
}
| null | null | null | null | true | null |
15637
| null |
Default
| null | null |
null |
{
"abstract": " To aid a variety of research studies, we propose TWIROLE, a hybrid model for\nrole-related user classification on Twitter, which detects male-related,\nfemale-related, and brand-related (i.e., organization or institution) users.\nTWIROLE leverages features from tweet contents, user profiles, and profile\nimages, and then applies our hybrid model to identify a user's role. To\nevaluate it, we used two existing large datasets about Twitter users, and\nconducted both intra- and inter-comparison experiments. TWIROLE outperforms\nexisting methods and obtains more balanced results over the several roles. We\nalso confirm that user names and profile images are good indicators for this\ntask. Our research extends prior work that does not consider brand-related\nusers, and is an aid to future evaluation efforts relative to investigations\nthat rely upon self-labeled datasets.\n",
"title": "A Hybrid Model for Role-related User Classification on Twitter"
}
| null | null | null | null | true | null |
15638
| null |
Default
| null | null |
null |
{
"abstract": " The protection of user privacy is an important concern in machine learning,\nas evidenced by the rolling out of the General Data Protection Regulation\n(GDPR) in the European Union (EU) in May 2018. The GDPR is designed to give\nusers more control over their personal data, which motivates us to explore\nmachine learning frameworks with data sharing without violating user privacy.\nTo meet this goal, in this paper, we propose a novel lossless\nprivacy-preserving tree-boosting system known as SecureBoost in the setting of\nfederated learning. This federated-learning system allows a learning process to\nbe jointly conducted over multiple parties with partially common user samples\nbut different feature sets, which corresponds to a vertically partitioned\nvirtual data set. An advantage of SecureBoost is that it provides the same\nlevel of accuracy as the non-privacy-preserving approach while at the same\ntime, reveal no information of each private data provider. We theoretically\nprove that the SecureBoost framework is as accurate as other non-federated\ngradient tree-boosting algorithms that bring the data into one place. In\naddition, along with a proof of security, we discuss what would be required to\nmake the protocols completely secure.\n",
"title": "SecureBoost: A Lossless Federated Learning Framework"
}
| null | null | null | null | true | null |
15639
| null |
Default
| null | null |
null |
{
"abstract": " We have studied the peculiarities of selective reflection from Rb vapor cell\nwith thickness $L <$ 70 nm, which is over an order of magnitude smaller than\nthe resonant wavelength for Rb atomic D$_1$ line $\\lambda$ = 795 nm. A huge\n($\\approx$ 240 MHz) red shift and spectral broadening of reflection signal is\nrecorded for $L =$ 40 nm caused by the atom-surface interaction. Also\ncompletely frequency resolved hyperfine Paschen-Back splitting of atomic\ntransitions to four components for $^{87}$Rb and six components for $^{85}$Rb\nis recorded in strong magnetic field ($B >$ 2 kG).\n",
"title": "Selective reflection from Rb layer with thickness below $λ$/12 and applications"
}
| null | null | null | null | true | null |
15640
| null |
Default
| null | null |
null |
{
"abstract": " The lambda-calculus is a peculiar computational model whose definition does\nnot come with a notion of machine. Unsurprisingly, implementations of the\nlambda-calculus have been studied for decades. Abstract machines are\nimplementations schema for fixed evaluation strategies that are a compromise\nbetween theory and practice: they are concrete enough to provide a notion of\nmachine and abstract enough to avoid the many intricacies of actual\nimplementations. There is an extensive literature about abstract machines for\nthe lambda-calculus, and yet-quite mysteriously-the efficiency of these\nmachines with respect to the strategy that they implement has almost never been\nstudied.\nThis paper provides an unusual introduction to abstract machines, based on\nthe complexity of their overhead with respect to the length of the implemented\nstrategies. It is conceived to be a tutorial, focusing on the case study of\nimplementing the weak head (call-by-name) strategy, and yet it is an original\nre-elaboration of known results. Moreover, some of the observation contained\nhere never appeared in print before.\n",
"title": "The Complexity of Abstract Machines"
}
| null | null | null | null | true | null |
15641
| null |
Default
| null | null |
null |
{
"abstract": " This paper can be viewed as a sequel to the author's long survey on the\nZimmer program \\cite{F11} published in 2011. The sequel focuses on recent rapid\nprogress on certain aspects of the program particularly concerning rigidity of\nAnosov actions and Zimmer's conjecture that there are no actions in low\ndimensions. Some emphasis is put on the surprising connections between these\ntwo different sets of developments and also on the key connections and ideas\nfor future research that arise from these works taken together.\n",
"title": "Recent progress in the Zimmer program"
}
| null | null | null | null | true | null |
15642
| null |
Default
| null | null |
null |
{
"abstract": " This paper proposes a new algorithm for recovery of belief network structure\nfrom data handling hidden variables. It consists essentially in an extension of\nthe CI algorithm of Spirtes et al. by restricting the number of conditional\ndependencies checked up to k variables and in an extension of the original CI\nby additional steps transforming so called partial including path graph into a\nbelief network. Its correctness is demonstrated.\n",
"title": "Restricted Causal Inference Algorithm"
}
| null | null | null | null | true | null |
15643
| null |
Default
| null | null |
null |
{
"abstract": " The mechanical properties of the cell depend crucially on the tension of its\ncytoskeleton, a biopolymer network that is put under stress by active motor\nproteins. While the fibrous nature of the network is known to strongly affect\nthe transmission of these forces to the cellular scale, our understanding of\nthis process remains incomplete. Here we investigate the transmission of forces\nthrough the network at the individual filament level, and show that active\nforces can be geometrically amplified as a transverse motor-generated force\nforce \"plucks\" the fiber and induces a nonlinear tension. In stiff and densely\nconnnected networks, this tension results in large network-wide tensile\nstresses that far exceed the expectation drawn from a linear elastic theory.\nThis amplification mechanism competes with a recently characterized\nnetwork-level amplification due to fiber buckling, suggesting that that fiber\nnetworks provide several distinct pathways for living systems to amplify their\nmolecular forces.\n",
"title": "Fiber plucking by molecular motors yields large emergent contractility in stiff biopolymer networks"
}
| null | null | null | null | true | null |
15644
| null |
Default
| null | null |
null |
{
"abstract": " The widespread adoption and dissemination of online news through social media\nsystems have been revolutionizing many segments of our society and ultimately\nour daily lives. In these systems, users can play a central role as they share\ncontent to their friends. Despite that, little is known about news spreaders in\nsocial media. In this paper, we provide the first of its kind in-depth\ncharacterization of news spreaders in social media. In particular, we\ninvestigate their demographics, what kind of content they share, and the\naudience they reach. Among our main findings, we show that males and white\nusers tend to be more active in terms of sharing news, biasing the news\naudience to the interests of these demographic groups. Our results also\nquantify differences in interests of news sharing across demographics, which\nhas implications for personalized news digests.\n",
"title": "Demographics of News Sharing in the U.S. Twittersphere"
}
| null | null | null | null | true | null |
15645
| null |
Default
| null | null |
null |
{
"abstract": " The interaction of (CH3-C5H4)Pt(CH3)3\n((methylcyclopentadienyl)trimethylplatinum)) molecules on fully and partially\nhydroxylated SiO2 surfaces, as well as the dynamics of this interaction were\ninvestigated using density functional theory (DFT) and finite temperature\nDFT-based molecular dynamics simulations. Fully and partially hydroxylated\nsurfaces represent substrates before and after electron beam treatment and this\nstudy examines the role of electron beam pretreatment on the substrates in the\ninitial stages of precursor dissociation and formation of Pt deposits. Our\nsimulations show that on fully hydroxylated surfaces or untreated surfaces, the\nprecursor molecules remain inactivated while we observe fragmentation of\n(CH3-C5H4)Pt(CH3)3 on partially hydroxylated surfaces. The behavior of\nprecursor molecules on the partially hydroxylated surfaces has been found to\ndepend on the initial orientation of the molecule and the distribution of\nsurface active sites. Based on the observations from the simulations and\navailable experiments, we discuss possible dissociation channels of the\nprecursor.\n",
"title": "Dynamics and fragmentation mechanism of (CH3-C5H4)Pt(CH3)3 on SiO2 Surfaces"
}
| null | null | null | null | true | null |
15646
| null |
Default
| null | null |
null |
{
"abstract": " We examine the H$\\beta$ Lick index in a sample of $\\sim 24000$ massive ($\\rm\nlog(M/M_{\\odot})>10.75$) and passive early-type galaxies extracted from SDSS at\nz<0.3, in order to assess the reliability of this index to constrain the epoch\nof formation and age evolution of these systems. We further investigate the\npossibility of exploiting this index as \"cosmic chronometer\", i.e. to derive\nthe Hubble parameter from its differential evolution with redshift, hence\nconstraining cosmological models independently of other probes. We find that\nthe H$\\beta$ strength increases with redshift as expected in passive evolution\nmodels, and shows at each redshift weaker values in more massive galaxies.\nHowever, a detailed comparison of the observed index with the predictions of\nstellar population synthesis models highlights a significant tension, with the\nobserved index being systematically lower than expected. By analyzing the\nstacked spectra, we find a weak [NII]$\\lambda6584$ emission line (not\ndetectable in the single spectra) which anti-correlates with the mass, that can\nbe interpreted as a hint of the presence of ionized gas. We estimated the\ncorrection of the H$\\beta$ index by the residual emission component exploiting\ndifferent approaches, but find it very uncertain and model-dependent. We\nconclude that, while the qualitative trends of the observed H$\\beta$-z\nrelations are consistent with the expected passive and downsizing scenario, the\npossible presence of ionized gas even in the most massive and passive galaxies\nprevents to use this index for a quantitative estimate of the age evolution and\nfor cosmological applications.\n",
"title": "On the robustness of the H$β$ Lick index as a cosmic clock in passive early-type galaxies"
}
| null | null |
[
"Physics"
] | null | true | null |
15647
| null |
Validated
| null | null |
null |
{
"abstract": " We present a collective coordinate approach to study the collective behaviour\nof a finite ensemble of $N$ stochastic Kuramoto oscillators using two degrees\nof freedom; one describing the shape dynamics of the oscillators and one\ndescribing their mean phase. Contrary to the thermodynamic limit $N\\to\\infty$\nin which the mean phase of the cluster of globally synchronized oscillators is\nconstant in time, the mean phase of a finite-size cluster experiences Brownian\ndiffusion with a variance proportional to $1/N$. This finite-size effect is\nquantitatively well captured by our collective coordinate approach.\n",
"title": "Finite-size effects in a stochastic Kuramoto model"
}
| null | null | null | null | true | null |
15648
| null |
Default
| null | null |
null |
{
"abstract": " The class of selfdecomposable distributions in free probability theory was\nintroduced by Barndorff-Nielsen and the third named author. It constitutes a\nfairly large subclass of the freely infinitely divisible distributions, but so\nfar specific examples have been limited to Wigner's semicircle distributions,\nthe free stable distributions, two kinds of free gamma distributions and a few\nother examples. In this paper, we prove that the (classical) normal\ndistributions are freely selfdecomposable. More generally it is established\nthat the Askey-Wimp-Kerov distribution $\\mu_c$ is freely selfdecomposable for\nany $c$ in $[-1,0]$. The main ingredient in the proof is a general\ncharacterization of the freely selfdecomposable distributions in terms of the\nderivative of their free cumulant transform.\n",
"title": "The normal distribution is freely selfdecomposable"
}
| null | null | null | null | true | null |
15649
| null |
Default
| null | null |
null |
{
"abstract": " In cloud storage systems, hot data is usually replicated over multiple nodes\nin order to accommodate simultaneous access by multiple users as well as\nincrease the fault tolerance of the system. Recent cloud storage research has\nproposed using availability codes, which is a special class of erasure codes,\nas a more storage efficient way to store hot data. These codes enable data\nrecovery from multiple, small disjoint groups of servers. The number of the\nrecovery groups is referred to as the availability and the size of each group\nas the locality of the code. Until now, we have very limited knowledge on how\ncode locality and availability affect data access time. Data download from\nthese systems involves multiple fork-join queues operating in-parallel, making\nthe analysis of access time a very challenging problem. In this paper, we\npresent an approximate analysis of data access time in storage systems that\nemploy simplex codes, which are an important and in certain sense optimal class\nof availability codes. We consider and compare three strategies in assigning\ndownload requests to servers; first one aggressively exploits the storage\navailability for faster download, second one implements only load balancing,\nand the last one employs storage availability only for hot data download\nwithout incurring any negative impact on the cold data download.\n",
"title": "Simplex Queues for Hot-Data Download"
}
| null | null | null | null | true | null |
15650
| null |
Default
| null | null |
null |
{
"abstract": " The binomial system is an electoral system unique in the world. It was used\nto elect the senators and deputies of Chile during 27 years, from the return of\ndemocracy in 1990 until 2017. In this paper we study the real voting power of\nthe different political parties in the Senate of Chile during the whole\nbinomial period. We not only consider the different legislative periods, but\nalso any party changes between one period and the next. The real voting power\nis measured by considering power indices from cooperative game theory, which\nare based on the capability of the political parties to form winning\ncoalitions. With this approach, we can do an analysis that goes beyond the\nsimple count of parliamentary seats.\n",
"title": "Voting power of political parties in the Senate of Chile during the whole binomial system period: 1990-2017"
}
| null | null | null | null | true | null |
15651
| null |
Default
| null | null |
null |
{
"abstract": " We prove a highly uniform stability or \"almost-near\" theorem for dual\nlattices of lattices $L \\subseteq \\Bbb R^n$. More precisely, we show that, for\na vector $x$ from the linear span of a lattice $L \\subseteq \\Bbb R^n$, subject\nto $\\lambda_1(L) \\ge \\lambda > 0$, to be $\\varepsilon$-close to some vector\nfrom the dual lattice $L'$ of $L$, it is enough that the inner products $u\\,x$\nare $\\delta$-close (with $\\delta < 1/3$) to some integers for all vectors $u\n\\in L$ satisfying $\\| u \\| \\le r$, where $r > 0$ depends on $n$, $\\lambda$,\n$\\delta$ and $\\varepsilon$, only. This generalizes an earlier analogous result\nproved for integral vector lattices by M. Mačaj and the second author. The\nproof is nonconstructive, using the ultraproduct construction and a slight\nportion of nonstandard analysis.\n",
"title": "A uniform stability principle for dual lattices"
}
| null | null | null | null | true | null |
15652
| null |
Default
| null | null |
null |
{
"abstract": " pandapower is a Python based, BSD-licensed power system analysis tool aimed\nat automation of static and quasi-static analysis and optimization of balanced\npower systems. It provides power flow, optimal power flow, state estimation,\ntopological graph searches and short circuit calculations according to IEC\n60909. pandapower includes a Newton-Raphson power flow solver formerly based on\nPYPOWER, which has been accelerated with just-in-time compilation. Additional\nenhancements to the solver include the capability to model constant current\nloads, grids with multiple reference nodes and a connectivity check. The\npandapower network model is based on electric elements, such as lines, two and\nthree-winding transformers or ideal switches. All elements can be defined with\nnameplate parameters and are internally processed with equivalent circuit\nmodels, which have been validated against industry standard software tools. The\ntabular data structure used to define networks is based on the Python library\npandas, which allows comfortable handling of input and output parameters. The\nimplementation in Python makes pandapower easy to use and allows comfortable\nextension with third-party libraries. pandapower has been successfully applied\nin several grid studies as well as for educational purposes. A comprehensive,\npublicly available case-study demonstrates a possible application of pandapower\nin an automated time series calculation.\n",
"title": "pandapower - an Open Source Python Tool for Convenient Modeling, Analysis and Optimization of Electric Power Systems"
}
| null | null | null | null | true | null |
15653
| null |
Default
| null | null |
null |
{
"abstract": " We consider light-induced binding and motion of dielectric microparticles in\nan optical waveguide that gives rise to a back-action effect such as light\ntransmission oscillating with time. Modeling the particles by dielectric slabs\nallows us to solve the problem analytically and obtain a rich variety of\ndynamical regimes both for Newtonian and damped motion. This variety is clearly\nreflected in temporal oscillations of the light transmission. The\ncharacteristic frequencies of the oscillations are within the ultrasound range\nof the order of $10^{5}$ Hz for micron size particles and injected power of the\norder of 100 mW. In addition, we consider driven by propagating light dynamics\nof a dielectric particle inside a Fabry-Perot resonator. These phenomena pave a\nway for optical driving and monitoring of motion of particles in waveguides and\nresonators.\n",
"title": "Temporal oscillations of light transmission through dielectric microparticles subjected to optically induced motion"
}
| null | null | null | null | true | null |
15654
| null |
Default
| null | null |
null |
{
"abstract": " We study rates of convergence in central limit theorems for the partial sum\nof squares of general Gaussian sequences, using tools from analysis on Wiener\nspace. No assumption of stationarity, asymptotically or otherwise, is made. The\nmain theoretical tool is the so-called Optimal Fourth Moment Theorem\n\\cite{NP2015}, which provides a sharp quantitative estimate of the total\nvariation distance on Wiener chaos to the normal law. The only assumptions made\non the sequence are the existence of an asymptotic variance, that a\nleast-squares-type estimator for this variance parameter has a bias and a\nvariance which can be controlled, and that the sequence's auto-correlation\nfunction, which may exhibit long memory, has a no-worse memory than that of\nfractional Brownian motion with Hurst parameter }$H<3/4$.{\\ \\ Our main result\nis explicit, exhibiting the trade-off between bias, variance, and memory. We\napply our result to study drift parameter estimation problems for subfractional\nOrnstein-Uhlenbeck and bifractional Ornstein-Uhlenbeck processes with\nfixed-time-step observations. These are processes which fail to be stationary\nor self-similar, but for which detailed calculations result in explicit\nformulas for the estimators' asymptotic normality.\n",
"title": "Berry-Esséen bounds for parameter estimation of general Gaussian processes"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
15655
| null |
Validated
| null | null |
null |
{
"abstract": " Anomaly detection aims to detect abnormal events by a model of normality. It\nplays an important role in many domains such as network intrusion detection,\ncriminal activity identity and so on. With the rapidly growing size of\naccessible training data and high computation capacities, deep learning based\nanomaly detection has become more and more popular. In this paper, a new\ndomain-based anomaly detection method based on generative adversarial networks\n(GAN) is proposed. Minimum likelihood regularization is proposed to make the\ngenerator produce more anomalies and prevent it from converging to normal data\ndistribution. Proper ensemble of anomaly scores is shown to improve the\nstability of discriminator effectively. The proposed method has achieved\nsignificant improvement than other anomaly detection methods on Cifar10 and UCI\ndatasets.\n",
"title": "Anomaly Detection via Minimum Likelihood Generative Adversarial Networks"
}
| null | null | null | null | true | null |
15656
| null |
Default
| null | null |
null |
{
"abstract": " A magnetic adatom chain, proximity coupled to a conventional superconductor\nwith spin-orbit coupling, exhibits locally an odd-parity, spin-triplet pairing\namplitude. We show that the singlet-triplet junction, thus formed, leads to a\nnet spin accumulation in the near vicinity of the chain. The accumulated spins\nare polarized along the direction of the local $\\mathbf{d}$-vector for triplet\npairing and generate an enhanced persistent current flowing around the chain.\nThe spin polarization and the \"supercurrent\" reverse their directions beyond a\ncritical exchange coupling strength at which the singlet superconducting order\nchanges its sign on the chain. The current is strongly enhanced in the\ntopological superconducting regime where Majorana bound states appear at the\nchain ends. The current and the spin profile offer alternative routes to\ncharacterize the topological superconducting state in adatom chains and\nislands.\n",
"title": "Supercurrent as a Probe for Topological Superconductivity in Magnetic Adatom Chains"
}
| null | null | null | null | true | null |
15657
| null |
Default
| null | null |
null |
{
"abstract": " Certain fibered hyperbolic 3-manifolds admit a $\\mathit{\\text{layered veering\ntriangulation}}$, which can be constructed algorithmically given the stable\nlamination of the monodromy. These triangulations were introduced by Agol in\n2011, and have been further studied by several others in the years since. We\nobtain experimental results which shed light on the combinatorial structure of\nveering triangulations, and its relation to certain topological invariants of\nthe underlying manifold.\n",
"title": "Experimental statistics of veering triangulations"
}
| null | null | null | null | true | null |
15658
| null |
Default
| null | null |
null |
{
"abstract": " Graph representations offer powerful and intuitive ways to describe data in a\nmultitude of application domains. Here, we consider stochastic processes\ngenerating graphs and propose a methodology for detecting changes in\nstationarity of such processes. The methodology is general and considers a\nprocess generating attributed graphs with a variable number of vertices/edges,\nwithout the need to assume one-to-one correspondence between vertices at\ndifferent time steps. The methodology acts by embedding every graph of the\nstream into a vector domain, where a conventional multivariate change detection\nprocedure can be easily applied. We ground the soundness of our proposal by\nproving several theoretical results. In addition, we provide a specific\nimplementation of the methodology and evaluate its effectiveness on several\ndetection problems involving attributed graphs representing biological\nmolecules and drawings. Experimental results are contrasted with respect to\nsuitable baseline methods, demonstrating the effectiveness of our approach.\n",
"title": "Concept Drift and Anomaly Detection in Graph Streams"
}
| null | null | null | null | true | null |
15659
| null |
Default
| null | null |
null |
{
"abstract": " Computational Fluid Dynamics (CFD) is a hugely important subject with\napplications in almost every engineering field, however, fluid simulations are\nextremely computationally and memory demanding. Towards this end, we present\nLat-Net, a method for compressing both the computation time and memory usage of\nLattice Boltzmann flow simulations using deep neural networks. Lat-Net employs\nconvolutional autoencoders and residual connections in a fully differentiable\nscheme to compress the state size of a simulation and learn the dynamics on\nthis compressed form. The result is a computationally and memory efficient\nneural network that can be iterated and queried to reproduce a fluid\nsimulation. We show that once Lat-Net is trained, it can generalize to large\ngrid sizes and complex geometries while maintaining accuracy. We also show that\nLat-Net is a general method for compressing other Lattice Boltzmann based\nsimulations such as Electromagnetism.\n",
"title": "Lat-Net: Compressing Lattice Boltzmann Flow Simulations using Deep Neural Networks"
}
| null | null |
[
"Physics",
"Statistics"
] | null | true | null |
15660
| null |
Validated
| null | null |
null |
{
"abstract": " While first-order optimization methods such as stochastic gradient descent\n(SGD) are popular in machine learning (ML), they come with well-known\ndeficiencies, including relatively-slow convergence, sensitivity to the\nsettings of hyper-parameters such as learning rate, stagnation at high training\nerrors, and difficulty in escaping flat regions and saddle points. These issues\nare particularly acute in highly non-convex settings such as those arising in\nneural networks. Motivated by this, there has been recent interest in\nsecond-order methods that aim to alleviate these shortcomings by capturing\ncurvature information. In this paper, we report detailed empirical evaluations\nof a class of Newton-type methods, namely sub-sampled variants of trust region\n(TR) and adaptive regularization with cubics (ARC) algorithms, for non-convex\nML problems. In doing so, we demonstrate that these methods not only can be\ncomputationally competitive with hand-tuned SGD with momentum, obtaining\ncomparable or better generalization performance, but also they are highly\nrobust to hyper-parameter settings. Further, in contrast to SGD with momentum,\nwe show that the manner in which these Newton-type methods employ curvature\ninformation allows them to seamlessly escape flat regions and saddle points.\n",
"title": "Second-Order Optimization for Non-Convex Machine Learning: An Empirical Study"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
15661
| null |
Validated
| null | null |
null |
{
"abstract": " We consider the stochastic shortest path (SSP) problem for succinct Markov\ndecision processes (MDPs), where the MDP consists of a set of variables, and a\nset of nondeterministic rules that update the variables. First, we show that\nseveral examples from the AI literature can be modeled as succinct MDPs. Then\nwe present computational approaches for upper and lower bounds for the SSP\nproblem: (a)~for computing upper bounds, our method is polynomial-time in the\nimplicit description of the MDP; (b)~for lower bounds, we present a\npolynomial-time (in the size of the implicit description) reduction to\nquadratic programming. Our approach is applicable even to infinite-state MDPs.\nFinally, we present experimental results to demonstrate the effectiveness of\nour approach on several classical examples from the AI literature.\n",
"title": "Computational Approaches for Stochastic Shortest Path on Succinct MDPs"
}
| null | null | null | null | true | null |
15662
| null |
Default
| null | null |
null |
{
"abstract": " We study existence and properties of one-dimensional edge domain walls in\nultrathin ferromagnetic films with uniaxial in-plane magnetic anisotropy. In\nthese materials, the magnetization vector is constrained to lie entirely in the\nfilm plane, with the preferred directions dictated by the magnetocrystalline\neasy axis. We consider magnetization profiles in the vicinity of a straight\nfilm edge oriented at an arbitrary angle with respect to the easy axis. To\nminimize the micromagnetic energy, these profiles form transition layers in\nwhich the magnetization vector rotates away from the direction of the easy axis\nto align with the film edge. We prove existence of edge domain walls as\nminimizers of the appropriate one-dimensional micromagnetic energy functional\nand show that they are classical solutions of the associated Euler-Lagrange\nequation with Dirichlet boundary condition at the edge. We also perform a\nnumerical study of these one-dimensional domain walls and uncover further\nproperties of these domain wall profiles.\n",
"title": "One-dimensional in-plane edge domain walls in ultrathin ferromagnetic films"
}
| null | null | null | null | true | null |
15663
| null |
Default
| null | null |
null |
{
"abstract": " A Turmit is a Turing machine that works over a two-dimensional grid, that is,\nan agent that moves, reads and writes symbols over the cells of the grid. Its\nstate is an arrow and, depending on the symbol that it reads, it turns to the\nleft or to the right, switching the symbol at the same time. Several symbols\nare admitted, and the rule is specified by the turning sense that the machine\nhas over each symbol. Turmites are a generalization of Langtons ant, and they\npresent very complex and diverse behaviors. We prove that any Turmite, except\nfor those whose rule does not depend on the symbol, can simulate any Turing\nMachine. We also prove the P-completeness of prediction their future behavior\nby explicitly giving a log-space reduction from the Topological Circuit Value\nProblem. A similar result was already established for Langtons ant; here we use\na similar technique but prove a stronger notion of simulation, and for a more\ngeneral family.\n",
"title": "Nontrivial Turmites are Turing-universal"
}
| null | null | null | null | true | null |
15664
| null |
Default
| null | null |
null |
{
"abstract": " Bitcoin and other cryptocurrencies have surged in popularity over the last\ndecade. Although Bitcoin does not claim to provide anonymity for its users, it\nenjoys a public perception of being a `privacy-preserving' financial system. In\nreality, cryptocurrencies publish users' entire transaction histories in\nplaintext, albeit under a pseudonym; this is required for transaction\nvalidation. Therefore, if a user's pseudonym can be linked to their human\nidentity, the privacy fallout can be significant. Recently, researchers have\ndemonstrated deanonymization attacks that exploit weaknesses in the Bitcoin\nnetwork's peer-to-peer (P2P) networking protocols. In particular, the P2P\nnetwork currently forwards content in a structured way that allows observers to\ndeanonymize users. In this work, we redesign the P2P network from first\nprinciples with the goal of providing strong, provable anonymity guarantees. We\npropose a simple networking policy called Dandelion, which achieves\nnearly-optimal anonymity guarantees at minimal cost to the network's utility.\nWe also provide a practical implementation of Dandelion.\n",
"title": "Dandelion: Redesigning the Bitcoin Network for Anonymity"
}
| null | null |
[
"Computer Science"
] | null | true | null |
15665
| null |
Validated
| null | null |
null |
{
"abstract": " Let $\\mathbb{K}$ be an infinite field. We prove that if a variety of\nanti-commutative $\\mathbb{K}$-algebras - not necessarily associative, where\n$xx=0$ is an identity - is locally algebraically cartesian closed, then it must\nbe a variety of Lie algebras over $\\mathbb{K}$. In particular,\n$\\mathsf{Lie}_{\\mathbb{K}}$ is the largest such. Thus, for a given variety of\nanti-commutative $\\mathbb{K}$-algebras, the Jacobi identity becomes equivalent\nto a categorical condition: it is an identity in~$\\mathcal{V}$ if and only if\n$\\mathcal{V}$ is a subvariety of a locally algebraically cartesian closed\nvariety of anti-commutative $\\mathbb{K}$-algebras. This is based on a result\nsaying that an algebraically coherent variety of anti-commutative\n$\\mathbb{K}$-algebras is either a variety of Lie algebras or a variety of\nanti-associative algebras over $\\mathbb{K}$.\n",
"title": "A characterisation of Lie algebras amongst anti-commutative algebras"
}
| null | null |
[
"Mathematics"
] | null | true | null |
15666
| null |
Validated
| null | null |
null |
{
"abstract": " We present the properties of a magneto-optical trap (MOT) of CaF molecules.\nWe study the process of loading the MOT from a decelerated buffer-gas-cooled\nbeam, and how best to slow this molecular beam in order to capture the most\nmolecules. We determine how the number of molecules, the photon scattering\nrate, the oscillation frequency, damping constant, temperature, cloud size and\nlifetime depend on the key parameters of the MOT, especially the intensity and\ndetuning of the main cooling laser. We compare our results to analytical and\nnumerical models, to the properties of standard atomic MOTs, and to MOTs of SrF\nmolecules. We load up to $2 \\times 10^4$ molecules, and measure a maximum\nscattering rate of $2.5 \\times 10^6$ s$^{-1}$ per molecule, a maximum\noscillation frequency of 100 Hz, a maximum damping constant of 500 s$^{-1}$,\nand a minimum MOT rms radius of 1.5 mm. A minimum temperature of 730 $\\mu$K is\nobtained by ramping down the laser intensity to low values. The lifetime,\ntypically about 100 ms, is consistent with a leak out of the cooling cycle with\na branching ratio of about $6 \\times 10^{-6}$. The MOT has a capture velocity\nof about 11 m/s.\n",
"title": "Characteristics of a magneto-optical trap of molecules"
}
| null | null | null | null | true | null |
15667
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we construct a non-autonomous version of the Hietarinta\nequation [Hietarinta J., J. Phys. A: Math. Gen. 37 (2004), L67-L73] and study\nits integrability properties. We show that this equation possess linear growth\nof the degrees of iterates, generalized symmetries depending on arbitrary\nfunctions, and that it is Darboux integrable. We use the first integrals to\nprovide a general solution of this equation. In particular we show that this\nequation is a sub-case of the non-autonomous $Q_{\\rm V}$ equation, and we\nprovide a non-autonomous Möbius transformation to another equation found in\n[Hietarinta J., J. Nonlinear Math. Phys. 12 (2005), suppl. 2, 223-230] and\nappearing also in Boll's classification [Boll R., Ph.D. Thesis, Technische\nUniversität Berlin, 2012].\n",
"title": "Reconstructing a Lattice Equation: a Non-Autonomous Approach to the Hietarinta Equation"
}
| null | null | null | null | true | null |
15668
| null |
Default
| null | null |
null |
{
"abstract": " Performance-critical machine learning models should be robust to input\nperturbations not seen during training. Adversarial training is a method for\nimproving a model's robustness to some perturbations by including them in the\ntraining process, but this tends to exacerbate other vulnerabilities of the\nmodel. The adversarial training framework has the effect of translating the\ndata with respect to the cost function, while weight decay has a scaling\neffect. Although weight decay could be considered a crude regularization\ntechnique, it appears superior to adversarial training as it remains stable\nover a broader range of regimes and reduces all generalization errors. Equipped\nwith these abstractions, we provide key baseline results and methodology for\ncharacterizing robustness. The two approaches can be combined to yield one\nsmall model that demonstrates good robustness to several white-box attacks\nassociated with different metrics.\n",
"title": "Adversarial Training Versus Weight Decay"
}
| null | null |
[
"Statistics"
] | null | true | null |
15669
| null |
Validated
| null | null |
null |
{
"abstract": " This paper presents a realization of the approach to spatial 3D stereo of\nvisualization of 3D images with use parallel Graphics processing unit (GPU).\nThe experiments of realization of synthesis of images of a 3D stage by a method\nof trace of beams on GPU with Compute Unified Device Architecture (CUDA) have\nshown that 60 % of the time is spent for the decision of a computing problem\napproximately, the major part of time (40 %) is spent for transfer of data\nbetween the central processing unit and GPU for computations and the\norganization process of visualization. The study of the influence of increase\nin the size of the GPU network at the speed of computations showed importance\nof the correct task of structure of formation of the parallel computer network\nand general mechanism of parallelization. Keywords: Volumetric 3D\nvisualization, stereo 3D visualization, ray tracing, parallel computing on GPU,\nCUDA\n",
"title": "Ray tracing method for stereo image synthesis using CUDA"
}
| null | null | null | null | true | null |
15670
| null |
Default
| null | null |
null |
{
"abstract": " Fractional quantum Hall-superconductor heterostructures may provide a\nplatform towards non-abelian topological modes beyond Majoranas. However their\nquantitative theoretical study remains extremely challenging. We propose and\nimplement a numerical setup for studying edge states of fractional quantum Hall\ndroplets with a superconducting instability. The fully gapped edges carry a\ntopological degree of freedom that can encode quantum information protected\nagainst local perturbations. We simulate such a system numerically using exact\ndiagonalization by restricting the calculation to the quasihole-subspace of a\n(time-reversal symmetric) bilayer fractional quantum Hall system of Laughlin\n$\\nu=1/3$ states. We show that the edge ground states are permuted by\nspin-dependent flux insertion and demonstrate their fractional $6\\pi$ Josephson\neffect, evidencing their topological nature and the Cooper pairing of\nfractionalized quasiparticles.\n",
"title": "Numerical investigation of gapped edge states in fractional quantum Hall-superconductor heterostructures"
}
| null | null |
[
"Physics"
] | null | true | null |
15671
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper we present the first results of a pilot experiment in the\ncapture and interpretation of multimodal signals of human experts engaged in\nsolving challenging chess problems. Our goal is to investigate the extent to\nwhich observations of eye-gaze, posture, emotion and other physiological\nsignals can be used to model the cognitive state of subjects, and to explore\nthe integration of multiple sensor modalities to improve the reliability of\ndetection of human displays of awareness and emotion. We observed chess players\nengaged in problems of increasing difficulty while recording their behavior.\nSuch recordings can be used to estimate a participant's awareness of the\ncurrent situation and to predict ability to respond effectively to challenging\nsituations. Results show that a multimodal approach is more accurate than a\nunimodal one. By combining body posture, visual attention and emotion, the\nmultimodal approach can reach up to 93% of accuracy when determining player's\nchess expertise while unimodal approach reaches 86%. Finally this experiment\nvalidates the use of our equipment as a general and reproducible tool for the\nstudy of participants engaged in screen-based interaction and/or problem\nsolving.\n",
"title": "Multimodal Observation and Interpretation of Subjects Engaged in Problem Solving"
}
| null | null | null | null | true | null |
15672
| null |
Default
| null | null |
null |
{
"abstract": " Encoder-decoder GANs architectures (e.g., BiGAN and ALI) seek to add an\ninference mechanism to the GANs setup, consisting of a small encoder deep net\nthat maps data-points to their succinct encodings. The intuition is that being\nforced to train an encoder alongside the usual generator forces the system to\nlearn meaningful mappings from the code to the data-point and vice-versa, which\nshould improve the learning of the target distribution and ameliorate\nmode-collapse. It should also yield meaningful codes that are useful as\nfeatures for downstream tasks. The current paper shows rigorously that even on\nreal-life distributions of images, the encode-decoder GAN training objectives\n(a) cannot prevent mode collapse; i.e. the objective can be near-optimal even\nwhen the generated distribution has low and finite support (b) cannot prevent\nlearning meaningless codes for data -- essentially white noise. Thus if\nencoder-decoder GANs do indeed work then it must be due to reasons as yet not\nunderstood, since the training objective can be low even for meaningless\nsolutions.\n",
"title": "Theoretical limitations of Encoder-Decoder GAN architectures"
}
| null | null | null | null | true | null |
15673
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we are concerned with the existence of the least energy\nsign-changing solutions for the following fractional Schrödinger-Poisson\nsystem: \\begin{align*}\n\\left\\{ \\begin{aligned} &(-\\Delta)^{s} u+V(x)u+\\lambda\\phi(x)u=f(x, u),\\quad\n&\\text{in}\\, \\ \\mathbb{R}^{3},\\\\ &(-\\Delta)^{t}\\phi=u^{2},& \\text{in}\\,\\\n\\mathbb{R}^{3}, \\end{aligned} \\right. \\end{align*} where $\\lambda\\in\n\\mathbb{R}^{+}$ is a parameter, $s, t\\in (0, 1)$ and $4s+2t>3$, $(-\\Delta)^{s}$\nstands for the fractional Laplacian. By constraint variational method and\nquantitative deformation lemma, we prove that the above problem has one least\nenergy sign-changing solution. Moreover, for any $\\lambda>0$, we show that the\nenergy of the least energy sign-changing solutions is strictly larger than two\ntimes the ground state energy.\nFinally, we consider $\\lambda$ as a parameter and study the convergence\nproperty of the least energy sign-changing solutions as $\\lambda\\searrow 0$.\n",
"title": "Ground state sign-changing solutions for a class of nonlinear fractional Schrödinger-Poisson system in $\\mathbb{R}^{3}$"
}
| null | null | null | null | true | null |
15674
| null |
Default
| null | null |
null |
{
"abstract": " We study the action of monads on categories equipped with several monoidal\nstructures. We identify the structure and conditions that guarantee that the\nhigher monoidal structure is inherited by the category of algebras over the\nmonad. Monoidal monads and comonoidal monads appear as the base cases in this\nhierarchy. Monads acting on duoidal categories constitute the next case. We\ncover the general case of $n$-monoidal categories and discuss several naturally\noccurring examples in which $n\\leq 3$.\n",
"title": "Monads on higher monoidal categories"
}
| null | null | null | null | true | null |
15675
| null |
Default
| null | null |
null |
{
"abstract": " Evacuation is one of the main disaster management solutions to reduce the\nimpact of man-made and natural threats on building occupants. To date, several\nmodern technologies and gamification concepts, e.g. immersive virtual reality\nand serious games, have been used to enhance building evacuation preparedness\nand effectiveness. Those tools have been used both to investigate human\nbehavior during building emergencies and to train building occupants on how to\ncope with building evacuations.\nAugmented Reality (AR) is novel technology that can enhance this process\nproviding building occupants with virtual contents to improve their evacuation\nperformance. This work aims at reviewing existing AR applications developed for\nbuilding evacuation. This review identifies the disasters and types of building\nthose tools have been applied for. Moreover, the application goals, hardware\nand evacuation stages affected by AR are also investigated in the review.\nFinally, this review aims at identifying the challenges to face for further\ndevelopment of AR evacuation tools.\n",
"title": "A Review of Augmented Reality Applications for Building Evacuation"
}
| null | null | null | null | true | null |
15676
| null |
Default
| null | null |
null |
{
"abstract": " The discovery of topological insulators has reformed modern materials\nscience, promising to be a platform for tabletop relativistic physics,\nelectronic transport without scattering, and stable quantum computation.\nTopological invariants are used to label distinct types of topological\ninsulators. But it is not generally known how many or which invariants can\nexist in any given crystalline material. Using a new and efficient counting\nalgorithm, we study the topological invariants that arise in time-reversal\nsymmetric crystals. This results in a unified picture that explains the\nrelations between all known topological invariants in these systems. It also\npredicts new topological phases and one entirely new topological invariant. We\npresent explicitly the classification of all two-dimensional crystalline\nfermionic materials, and give a straightforward procedure for finding the\nanalogous result in any three-dimensional structure. Our study represents a\nsingle, intuitive physical picture applicable to all topological invariants in\nreal materials, with crystal symmetries.\n",
"title": "Topology in time-reversal symmetric crystals"
}
| null | null | null | null | true | null |
15677
| null |
Default
| null | null |
null |
{
"abstract": " Deep convolutional neural network (CNN) inference requires significant amount\nof memory and computation, which limits its deployment on embedded devices. To\nalleviate these problems to some extent, prior research utilize low precision\nfixed-point numbers to represent the CNN weights and activations. However, the\nminimum required data precision of fixed-point weights varies across different\nnetworks and also across different layers of the same network. In this work, we\npropose using floating-point numbers for representing the weights and\nfixed-point numbers for representing the activations. We show that using\nfloating-point representation for weights is more efficient than fixed-point\nrepresentation for the same bit-width and demonstrate it on popular large-scale\nCNNs such as AlexNet, SqueezeNet, GoogLeNet and VGG-16. We also show that such\na representation scheme enables compact hardware multiply-and-accumulate (MAC)\nunit design. Experimental results show that the proposed scheme reduces the\nweight storage by up to 36% and power consumption of the hardware multiplier by\nup to 50%.\n",
"title": "Deep Convolutional Neural Network Inference with Floating-point Weights and Fixed-point Activations"
}
| null | null | null | null | true | null |
15678
| null |
Default
| null | null |
null |
{
"abstract": " The formation of large voids in the Cosmic Web from the initial adiabatic\ncosmological perturbations of space-time metric, density and velocity of matter\nis investigated in cosmological model with the dynamical dark energy\naccelerating expansion of the Universe. It is shown that the negative density\nperturbations with the initial radius of about 50 Mpc in comoving to the\ncosmological background coordinates and the amplitude corresponding to the\nr.m.s. temperature fluctuations of the cosmic microwave background lead to the\nformation of voids with the density contrast up to $-$0.9, maximal peculiar\nvelocity about 400 km/s and the radius close to the initial one. An important\nfeature of voids formation from the analyzed initial amplitudes and profiles is\nestablishing the surrounding overdensity shell. We have shown that the ratio of\nthe peculiar velocity in units of the Hubble flow to the density contrast in\nthe central part of a void does not depend or weakly depends on the distance\nfrom the center of the void. It is also shown that this ratio is sensitive to\nthe values of dark energy parameters and can be used to find them based on the\nobservational data on mass density and peculiar velocities of galaxies in the\nvoids.\n",
"title": "Voids in the Cosmic Web as a probe of dark energy"
}
| null | null | null | null | true | null |
15679
| null |
Default
| null | null |
null |
{
"abstract": " The stability of sequence replication was crucial for the emergence of\nmolecular evolution and early life. Exponential replication with a first-order\ngrowth dynamics show inherent instabilities such as the error catastrophe and\nthe dominance by the fastest replicators. This favors less structured and short\nsequences. The theoretical concept of hypercycles has been proposed to solve\nthese problems. Their higher-order growth kinetics leads to frequency-dependent\nselection and stabilizes the replication of majority molecules. However, many\nimplementations of hypercycles are unstable or require special sequences with\ncatalytic activity. Here, we demonstrate the spontaneous emergence of\nhigher-order cooperative replication from a network of simple ligation chain\nreactions (LCR). We performed long-term LCR experiments from a mixture of\nsequences under molecule degrading conditions with a ligase protein. At the\nchosen temperature cycling, a network of positive feedback loops arose from\nboth the cooperative ligation of matching sequences and the emerging increase\nin sequence length. It generated higher-order replication with\nfrequency-dependent selection. The experiments matched a complete simulation\nusing experimentally determined ligation rates and the hypercycle mechanism was\nalso confirmed by abstracted modeling. Since templated ligation is a most basic\nreaction of oligonucleotides, the described mechanism could have been\nimplemented under microthermal convection on early Earth.\n",
"title": "Templated ligation can create a hypercycle replication network"
}
| null | null |
[
"Quantitative Biology"
] | null | true | null |
15680
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper we obtain the variational characterization of Hardy space $H^p$\nfor $p\\in(\\frac n{n+1},1]$ and get estimates for the oscillation operator and\nthe $\\lambda$-jump operator associated with approximate identities acting on\n$H^p$ for $p\\in(\\frac n{n+1},1]$. Moreover, we give counterexamples to show\nthat the oscillation and $\\lambda$-jump associated with some approximate\nidentity can not be used to characterize $H^p$ for $p\\in(\\frac n{n+1},1]$.\n",
"title": "Variational characterization of H^p"
}
| null | null | null | null | true | null |
15681
| null |
Default
| null | null |
null |
{
"abstract": " Recent work has demonstrated that neural networks are vulnerable to\nadversarial examples, i.e., inputs that are almost indistinguishable from\nnatural data and yet classified incorrectly by the network. In fact, some of\nthe latest findings suggest that the existence of adversarial attacks may be an\ninherent weakness of deep learning models. To address this problem, we study\nthe adversarial robustness of neural networks through the lens of robust\noptimization. This approach provides us with a broad and unifying view on much\nof the prior work on this topic. Its principled nature also enables us to\nidentify methods for both training and attacking neural networks that are\nreliable and, in a certain sense, universal. In particular, they specify a\nconcrete security guarantee that would protect against any adversary. These\nmethods let us train networks with significantly improved resistance to a wide\nrange of adversarial attacks. They also suggest the notion of security against\na first-order adversary as a natural and broad security guarantee. We believe\nthat robustness against such well-defined classes of adversaries is an\nimportant stepping stone towards fully resistant deep learning models.\n",
"title": "Towards Deep Learning Models Resistant to Adversarial Attacks"
}
| null | null | null | null | true | null |
15682
| null |
Default
| null | null |
null |
{
"abstract": " We present a first internal delensing of CMB maps, both in temperature and\npolarization, using the public foreground-cleaned (SMICA) Planck 2015 maps.\nAfter forming quadratic estimates of the lensing potential, we use the\ncorresponding displacement field to undo the lensing on the same data. We build\ndifferences of the delensed spectra to the original data spectra specifically\nto look for delensing signatures. After taking into account reconstruction\nnoise biases in the delensed spectra, we find an expected sharpening of the\npower spectrum acoustic peaks with a delensing efficiency of $29\\,\\%$ ($TT$)\n$25\\,\\%$ ($TE$) and $22\\,\\%$ ($EE$). The detection significance of the\ndelensing effects is very high in all spectra: $12\\,\\sigma$ in $EE$\npolarization; $18\\,\\sigma$ in $TE$; and $20\\,\\sigma$ in $TT$. The null\nhypothesis of no lensing in the maps is rejected at $26\\,\\sigma$. While direct\ndetection of the power in lensing $B$-modes themselves is not possible at high\nsignificance at Planck noise levels, we do detect (at $4.5\\,\\sigma$ under the\nnull hypothesis) delensing effects in the $B$-mode map, with $7\\,\\%$ reduction\nin lensing power. Our results provide a first demonstration of polarization\ndelensing, and generally of internal CMB delensing, and stand in agreement with\nthe baseline $\\Lambda$CDM Planck 2015 cosmology expectations.\n",
"title": "Internal delensing of Planck CMB temperature and polarization"
}
| null | null | null | null | true | null |
15683
| null |
Default
| null | null |
null |
{
"abstract": " By introducing programmability, automated verification, and innovative\ndebugging tools, Software-Defined Networks (SDNs) are poised to meet the\nincreasingly stringent dependability requirements of today's communication\nnetworks. However, the design of fault-tolerant SDNs remains an open challenge.\nThis paper considers the design of dependable SDNs through the lenses of\nself-stabilization - a very strong notion of fault-tolerance. In particular, we\ndevelop algorithms for an in-band and distributed control plane for SDNs,\ncalled Renaissance, which tolerate a wide range of (concurrent) controller,\nlink, and communication failures. Our self-stabilizing algorithms ensure that\nafter the occurrence of an arbitrary combination of failures, (i) every\nnon-faulty SDN controller can eventually reach any switch in the network within\na bounded communication delay (in the presence of a bounded number of\nconcurrent failures) and (ii) every switch is managed by at least one\nnon-faulty controller. We evaluate Renaissance through a rigorous worst-case\nanalysis as well as a prototype implementation (based on OVS and Floodlight),\nand we report on our experiments using Mininet.\n",
"title": "Renaissance: Self-Stabilizing Distributed SDN Control Plane"
}
| null | null | null | null | true | null |
15684
| null |
Default
| null | null |
null |
{
"abstract": " In today's cyber-enabled smart grids, high penetration of uncertain\nrenewables, purposeful manipulation of meter readings, and the need for\nwide-area situational awareness, call for fast, accurate, and robust power\nsystem state estimation. The least-absolute-value (LAV) estimator is known for\nits robustness relative to the weighted least-squares (WLS) one. However, due\nto nonconvexity and nonsmoothness, existing LAV solvers based on linear\nprogramming are typically slow, hence inadequate for real-time system\nmonitoring. This paper develops two novel algorithms for efficient LAV\nestimation, which draw from recent advances in composite optimization. The\nfirst is a deterministic linear proximal scheme that handles a sequence of\nconvex quadratic problems, each efficiently solvable either via off-the-shelf\nalgorithms or through the alternating direction method of multipliers.\nLeveraging the sparse connectivity inherent to power networks, the second\nscheme is stochastic, and updates only \\emph{a few} entries of the complex\nvoltage state vector per iteration. In particular, when voltage magnitude and\n(re)active power flow measurements are used only, this number reduces to one or\ntwo, \\emph{regardless of} the number of buses in the network. This\ncomputational complexity evidently scales well to large-size power systems.\nFurthermore, by carefully \\emph{mini-batching} the voltage and power flow\nmeasurements, accelerated implementation of the stochastic iterations becomes\npossible. The developed algorithms are numerically evaluated using a variety of\nbenchmark power networks. Simulated tests corroborate that improved robustness\ncan be attained at comparable or markedly reduced computation times for medium-\nor large-size networks relative to the \"workhorse\" WLS-based Gauss-Newton\niterations.\n",
"title": "Robust and Scalable Power System State Estimation via Composite Optimization"
}
| null | null | null | null | true | null |
15685
| null |
Default
| null | null |
null |
{
"abstract": " In this article, we provide a new algorithm for solving constraint\nsatisfaction problems over templates with few subpowers, by reducing the\nproblem to the combination of solvability of a polynomial number of systems of\nlinear equations over finite fields and reductions via absorbing subuniverses.\n",
"title": "A new algorithm for constraint satisfaction problems with few subpowers templates"
}
| null | null | null | null | true | null |
15686
| null |
Default
| null | null |
null |
{
"abstract": " The geodetic VLBI technique is capable of measuring the Sun's gravity light\ndeflection from distant radio sources around the whole sky. This light\ndeflection is equivalent to the conventional gravitational delay used for the\nreduction of geodetic VLBI data. While numerous tests based on a global set of\nVLBI data have shown that the parameter 'gamma' of the post-Newtonian\napproximation is equal to unity with a precision of about 0.02 percent, more\ndetailed analysis reveals some systematic deviations depending on the angular\nelongation from the Sun. In this paper a limited set of VLBI observations near\nthe Sun were adjusted to obtain the estimate of the parameter 'gamma' free of\nthe elongation angle impact. The parameter 'gamma' is still found to be close\nto unity with precision of 0.06 percent, two subsets of VLBI data measured at\nshort and long baselines produce some statistical inconsistency.\n",
"title": "Testing of General Relativity with Geodetic VLBI"
}
| null | null | null | null | true | null |
15687
| null |
Default
| null | null |
null |
{
"abstract": " An unbiased estimator for the ellipticity of an object in a noisy image is\ngiven in terms of the image moments. Three assumptions are made: i) the pixel\nnoise is normally distributed, although with arbitrary covariance matrix, ii)\nthe image moments are taken about a fixed centre, and iii) the point-spread\nfunction is known. The relevant combinations of image moments are then jointly\nnormal and their covariance matrix can be computed. A particular estimator for\nthe ratio of the means of jointly normal variates is constructed and used to\nprovide the unbiased estimator for the ellipticity. Furthermore, an unbiased\nestimate of the covariance of the new estimator is also given.\n",
"title": "An unbiased estimator for the ellipticity from image moments"
}
| null | null | null | null | true | null |
15688
| null |
Default
| null | null |
null |
{
"abstract": " There have been several spectral bounds for the percolation transition in\nnetworks, using spectrum of matrices associated with the network such as the\nadjacency matrix and the non-backtracking matrix. However they are far from\nbeing tight when the network is sparse and displays clustering or transitivity,\nwhich is represented by existence of short loops e.g. triangles. In this work,\nfor the bond percolation, we first propose a message passing algorithm for\ncalculating size of percolating clusters considering effects of triangles, then\nrelate the percolation transition to the leading eigenvalue of a matrix that we\nname the triangle-non-backtracking matrix, by analyzing stability of the\nmessage passing equations. We establish that our method gives a tighter\nlower-bound to the bond percolation transition than previous spectral bounds,\nand it becomes exact for an infinite network with no loops longer than 3. We\nevaluate numerically our methods on synthetic and real-world networks, and\ndiscuss further generalizations of our approach to include higher-order\nsub-structures.\n",
"title": "Spectral estimation of the percolation transition in clustered networks"
}
| null | null | null | null | true | null |
15689
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we extend the works of Tancer and of Malgouyres and Francés,\nshowing that $(d,k)$-collapsibility is NP-complete for $d\\geq k+2$ except\n$(2,0)$. By $(d,k)$-collapsibility we mean the following problem: determine\nwhether a given $d$-dimensional simplicial complex can be collapsed to some\n$k$-dimensional subcomplex. The question of establishing the complexity status\nof $(d,k)$-collapsibility was asked by Tancer, who proved NP-completeness of\n$(d,0)$ and $(d,1)$-collapsibility (for $d\\geq 3$). Our extended result,\ntogether with the known polynomial-time algorithms for $(2,0)$ and $d=k+1$,\nanswers the question completely.\n",
"title": "Collapsibility to a subcomplex of a given dimension is NP-complete"
}
| null | null | null | null | true | null |
15690
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we present our approach to solve a physics-based reinforcement\nlearning challenge \"Learning to Run\" with objective to train\nphysiologically-based human model to navigate a complex obstacle course as\nquickly as possible. The environment is computationally expensive, has a\nhigh-dimensional continuous action space and is stochastic. We benchmark state\nof the art policy-gradient methods and test several improvements, such as layer\nnormalization, parameter noise, action and state reflecting, to stabilize\ntraining and improve its sample-efficiency. We found that the Deep\nDeterministic Policy Gradient method is the most efficient method for this\nenvironment and the improvements we have introduced help to stabilize training.\nLearned models are able to generalize to new physical scenarios, e.g. different\nobstacle courses.\n",
"title": "Run, skeleton, run: skeletal model in a physics-based simulation"
}
| null | null | null | null | true | null |
15691
| null |
Default
| null | null |
null |
{
"abstract": " Acquisition of labeled training samples for affective computing is usually\ncostly and time-consuming, as affects are intrinsically subjective, subtle and\nuncertain, and hence multiple human assessors are needed to evaluate each\naffective sample. Particularly, for affect estimation in the 3D space of\nvalence, arousal and dominance, each assessor has to perform the evaluations in\nthree dimensions, which makes the labeling problem even more challenging. Many\nsophisticated machine learning approaches have been proposed to reduce the data\nlabeling requirement in various other domains, but so far few have considered\naffective computing. This paper proposes two multi-task active learning for\nregression approaches, which select the most beneficial samples to label, by\nconsidering the three affect primitives simultaneously. Experimental results on\nthe VAM corpus demonstrated that our optimal sample selection approaches can\nresult in better estimation performance than random selection and several\ntraditional single-task active learning approaches. Thus, they can help\nalleviate the data labeling problem in affective computing, i.e., better\nestimation performance can be obtained from fewer labeling queries.\n",
"title": "Affect Estimation in 3D Space Using Multi-Task Active Learning for Regression"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
15692
| null |
Validated
| null | null |
null |
{
"abstract": " In this paper, a sparse Markov decision process (MDP) with novel causal\nsparse Tsallis entropy regularization is proposed.The proposed policy\nregularization induces a sparse and multi-modal optimal policy distribution of\na sparse MDP. The full mathematical analysis of the proposed sparse MDP is\nprovided.We first analyze the optimality condition of a sparse MDP. Then, we\npropose a sparse value iteration method which solves a sparse MDP and then\nprove the convergence and optimality of sparse value iteration using the Banach\nfixed point theorem. The proposed sparse MDP is compared to soft MDPs which\nutilize causal entropy regularization. We show that the performance error of a\nsparse MDP has a constant bound, while the error of a soft MDP increases\nlogarithmically with respect to the number of actions, where this performance\nerror is caused by the introduced regularization term. In experiments, we apply\nsparse MDPs to reinforcement learning problems. The proposed method outperforms\nexisting methods in terms of the convergence speed and performance.\n",
"title": "Sparse Markov Decision Processes with Causal Sparse Tsallis Entropy Regularization for Reinforcement Learning"
}
| null | null | null | null | true | null |
15693
| null |
Default
| null | null |
null |
{
"abstract": " The dust-forming nova V2676 Oph is unique in that it was the first nova to\nprovide evidence of C_2 and CN molecules during its near-maximum phase and\nevidence of CO molecules during its early decline phase. Observations of this\nnova have revealed the slow evolution of its lightcurves and have also shown\nlow isotopic ratios of carbon (12C/13C) and nitrogen (14N/15N) in its nova\nenvelope. These behaviors indicate that the white dwarf (WD) star hosting V2676\nOph is a CO-rich WD rather than an ONe-rich WD (typically larger in mass than\nthe former). We performed mid-infrared spectroscopic and photometric\nobservations of V2676 Oph in 2013 and 2014 (respectively 452 and 782 days after\nits discovery). No significant [Ne II] emission at 12.8 micron was detected at\neither epoch. These provided evidence for a CO-rich WD star hosting V2676 Oph.\nBoth carbon-rich and oxygen-rich grains were detected in addition to an\nunidentified infrared feature at 11.4 micron originating from polycyclic\naromatic hydrocarbon molecules or hydrogenated amorphous carbon grains in the\nenvelope of V2676 Oph.\n",
"title": "Mid-infrared Spectroscopic Observations of the Dust-forming Classical Nova V2676 Oph"
}
| null | null | null | null | true | null |
15694
| null |
Default
| null | null |
null |
{
"abstract": " This article consists of two parts. In Part 1, we present a formulation of\ntwo-dimensional topological quantum field theories in terms of a functor from a\ncategory of Ribbon graphs to the endofuntor category of a monoidal category.\nThe key point is that the category of ribbon graphs produces all Frobenius\nobjects. Necessary backgrounds from Frobenius algebras, topological quantum\nfield theories, and cohomological field theories are reviewed. A result on\nFrobenius algebra twisted topological recursion is included at the end of Part\n1.\nIn Part 2, we explain a geometric theory of quantum curves. The focus is\nplaced on the process of quantization as a passage from families of Hitchin\nspectral curves to families of opers. To make the presentation simpler, we\nunfold the story using SL_2(\\mathbb{C})-opers and rank 2 Higgs bundles defined\non a compact Riemann surface $C$ of genus greater than $1$. In this case,\nquantum curves, opers, and projective structures in $C$ all become the same\nnotion. Background materials on projective coordinate systems, Higgs bundles,\nopers, and non-Abelian Hodge correspondence are explained.\n",
"title": "An invitation to 2D TQFT and quantization of Hitchin spectral curves"
}
| null | null | null | null | true | null |
15695
| null |
Default
| null | null |
null |
{
"abstract": " Data stream mining problem has caused widely concerns in the area of machine\nlearning and data mining. In some recent studies, ensemble classification has\nbeen widely used in concept drift detection, however, most of them regard\nclassification accuracy as a criterion for judging whether concept drift\nhappening or not. Information entropy is an important and effective method for\nmeasuring uncertainty. Based on the information entropy theory, a new algorithm\nusing information entropy to evaluate a classification result is developed. It\nuses ensemble classification techniques, and the weight of each classifier is\ndecided through the entropy of the result produced by an ensemble classifiers\nsystem. When the concept in data streams changing, the classifiers' weight\nbelow a threshold value will be abandoned to adapt to a new concept in one\ntime. In the experimental analysis section, six databases and four proposed\nalgorithms are executed. The results show that the proposed method can not only\nhandle concept drift effectively, but also have a better classification\naccuracy and time performance than the contrastive algorithms.\n",
"title": "An Ensemble Classification Algorithm Based on Information Entropy for Data Streams"
}
| null | null |
[
"Computer Science"
] | null | true | null |
15696
| null |
Validated
| null | null |
null |
{
"abstract": " We investigate the low-energy scaling behavior of an interacting 3D Weyl\nsemimetal in the presence of disorder. In order to achieve a renormalization\ngroup analysis of the theory, we focus on the effects of a\nshort-ranged-correlated disorder potential, checking nevertheless that this\nchoice is not essential to locate the different phases of the Weyl semimetal.\nWe show that there is a line of fixed-points in the renormalization group flow\nof the interacting theory, corresponding to the disorder-driven transition to a\ndiffusive metal phase. Along that boundary, the critical disorder strength\nundergoes a strong increase with respect to the noninteracting theory, as a\nconsequence of the unconventional screening of the Coulomb and disorder-induced\ninteractions. A complementary resolution of the Schwinger-Dyson equations\nallows us to determine the full phase diagram of the system, showing the\nprevalence of a renormalized semimetallic phase in the regime of intermediate\ninteraction strength, and adjacent to the non-Fermi liquid phase characteristic\nof the strong interaction regime of 3D Weyl semimetals.\n",
"title": "Competition between disorder and interaction effects in 3D Weyl semimetals"
}
| null | null | null | null | true | null |
15697
| null |
Default
| null | null |
null |
{
"abstract": " Positively (resp. negatively) associated point processes are a class of point\nprocesses that induce attraction (resp. inhibition) between the points. As an\nimportant example, determinantal point processes (DPPs) are negatively\nassociated. We prove $\\alpha$-mixing properties for associated spatial point\nprocesses by controlling their $\\alpha$-coefficients in terms of the first two\nintensity functions. A central limit theorem for functionals of associated\npoint processes is deduced, using both the association and the $\\alpha$-mixing\nproperties. We discuss in detail the case of DPPs, for which we obtain the\nlimiting distribution of sums, over subsets of close enough points of the\nprocess, of any bounded function of the DPP. As an application, we get the\nasymptotic properties of the parametric two-step estimator of some\ninhomogeneous DPPs.\n",
"title": "Mixing properties and central limit theorem for associated point processes"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
15698
| null |
Validated
| null | null |
null |
{
"abstract": " With the increasing abundance of 'digital footprints' left by human\ninteractions in online environments, e.g., social media and app use, the\nability to model complex human behavior has become increasingly possible. Many\napproaches have been proposed, however, most previous model frameworks are\nfairly restrictive. We introduce a new social modeling approach that enables\nthe creation of models directly from data with minimal a priori restrictions on\nthe model class. In particular, we infer the minimally complex, maximally\npredictive representation of an individual's behavior when viewed in isolation\nand as driven by a social input. We then apply this framework to a\nheterogeneous catalog of human behavior collected from fifteen thousand users\non the microblogging platform Twitter. The models allow us to describe how a\nuser processes their past behavior and their social inputs. Despite the\ndiversity of observed user behavior, most models inferred fall into a small\nsubclass of all possible finite-state processes. Thus, our work demonstrates\nthat user behavior, while quite complex, belies simple underlying computational\nstructures.\n",
"title": "Computational landscape of user behavior on social media"
}
| null | null | null | null | true | null |
15699
| null |
Default
| null | null |
null |
{
"abstract": " We present a finite difference time domain (FDTD) model for computation of A\nline scans in time domain optical coherence tomography (OCT). By simulating\nonly the end of the two arms of the interferometer and computing the\ninterference signal in post processing, it is possible to reduce the\ncomputational time required by the simulations and, thus, to simulate much\nbigger environments. Moreover, it is possible to simulate successive A lines\nand thus obtaining a cross section of the sample considered. In this paper we\npresent the model applied to two different samples: a glass rod filled with\nwater-sucrose solution at different concentrations and a peripheral nerve. This\nwork demonstrates the feasibility of using OCT for non-invasive, direct optical\nmonitoring of peripheral nerve activity, which is a long-sought goal of\nneuroscience.\n",
"title": "Simulating optical coherence tomography for observing nerve activity: a finite difference time domain bi-dimensional model"
}
| null | null | null | null | true | null |
15700
| null |
Default
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.