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": " Many modern clustering methods scale well to a large number of data items, N,\nbut not to a large number of clusters, K. This paper introduces PERCH, a new\nnon-greedy algorithm for online hierarchical clustering that scales to both\nmassive N and K--a problem setting we term extreme clustering. Our algorithm\nefficiently routes new data points to the leaves of an incrementally-built\ntree. Motivated by the desire for both accuracy and speed, our approach\nperforms tree rotations for the sake of enhancing subtree purity and\nencouraging balancedness. We prove that, under a natural separability\nassumption, our non-greedy algorithm will produce trees with perfect dendrogram\npurity regardless of online data arrival order. Our experiments demonstrate\nthat PERCH constructs more accurate trees than other tree-building clustering\nalgorithms and scales well with both N and K, achieving a higher quality\nclustering than the strongest flat clustering competitor in nearly half the\ntime.\n",
"title": "An Online Hierarchical Algorithm for Extreme Clustering"
}
| null | null | null | null | true | null |
12901
| null |
Default
| null | null |
null |
{
"abstract": " We review some recent results on geometric equations on Lorentzian manifolds\nsuch as the wave and Dirac equations. This includes well-posedness and\nstability for various initial value problems, as well as results on the\nstructure of these equations on black-hole spacetimes (in particular, on the\nKerr solution), the index theorem for hyperbolic Dirac operators and properties\nof the class of Green-hyperbolic operators.\n",
"title": "Wave and Dirac equations on manifolds"
}
| null | null | null | null | true | null |
12902
| null |
Default
| null | null |
null |
{
"abstract": " We study the recombination process of three atoms scattering into an atom and\ndiatomic molecule in heteronuclear mixtures of ultracold atomic gases with\nlarge and positive interspecies scattering length at finite temperature. We\ncalculate the temperature dependence of the three-body recombination rates by\nextracting universal scaling functions that parametrize the energy dependence\nof the scattering matrix. We compare our results to experimental data for the\n40K-87Rb mixture and make a prediction for 6Li-87Rb. We find that contributions\nfrom higher partial wave channels significantly impact the total rate and, in\nsystems with particularly large mass imbalance, can even obliterate the\nrecombination minima associated with the Efimov effect.\n",
"title": "The Efimov effect for heteronuclear three-body systems at positive scattering length and finite temperature"
}
| null | null | null | null | true | null |
12903
| null |
Default
| null | null |
null |
{
"abstract": " This paper, the second of a two-part series, presents a method for mean-field\nfeedback stabilization of a swarm of agents on a finite state space whose time\nevolution is modeled as a continuous time Markov chain (CTMC). The resulting\n(mean-field) control problem is that of controlling a nonlinear system with\ndesired global stability properties. We first prove that any probability\ndistribution with a strongly connected support can be stabilized using\ntime-invariant inputs. Secondly, we show the asymptotic controllability of all\npossible probability distributions, including distributions that assign zero\ndensity to some states and which do not necessarily have a strongly connected\nsupport. Lastly, we demonstrate that there always exists a globally\nasymptotically stabilizing decentralized density feedback law with the\nadditional property that the control inputs are zero at equilibrium, whenever\nthe graph is strongly connected and bidirected. Then the problem of\nsynthesizing closed-loop polynomial feedback is framed as a optimization\nproblem using state-of-the-art sum-of-squares optimization tools. The\noptimization problem searches for polynomial feedback laws that make the\ncandidate Lyapunov function a stability certificate for the resulting\nclosed-loop system. Our methodology is tested for two cases on a five vertex\ngraph, and the stabilization properties of the constructed control laws are\nvalidated with numerical simulations of the corresponding system of ordinary\ndifferential equations.\n",
"title": "Mean-Field Controllability and Decentralized Stabilization of Markov Chains, Part II: Asymptotic Controllability and Polynomial Feedbacks"
}
| null | null |
[
"Computer Science",
"Mathematics"
] | null | true | null |
12904
| null |
Validated
| null | null |
null |
{
"abstract": " A minimum in stellar velocity dispersion is often observed in the central\nregions of disc galaxies. To investigate the origin of this feature, known as a\n{\\sigma}-drop, we analyse the stellar kinematics of a high-resolution N-body +\nsmooth particle hydrodynamical simulation, which models the secular evolution\nof an unbarred disc galaxy. We compared the intrinsic mass-weighted kinematics\nto the recovered luminosity-weighted ones. The latter were obtained by\nanalysing synthetic spectra produced by a new code, SYNTRA, that generates\nsynthetic spectra by assigning a stellar population synthesis model to each\nstar particle based on its age and metallicity. The kinematics were derived\nfrom the synthetic spectra as in real spectra to mimic the kinematic analysis\nof real galaxies. We found that the recovered luminosity-weighted kinematics in\nthe centre of the simulated galaxy are biased to higher rotation velocities and\nlower velocity dispersions due to the presence of young stars in a thin and\nkinematically cool disc, and are ultimately responsible for the {\\sigma}-drop.\n",
"title": "The kinematics of σ-drop bulges from spectral synthesis modelling of a hydrodynamical simulation"
}
| null | null | null | null | true | null |
12905
| null |
Default
| null | null |
null |
{
"abstract": " Recommendation systems are recognised as being hugely important in industry,\nand the area is now well understood. At News UK, there is a requirement to be\nable to quickly generate recommendations for users on news items as they are\npublished. However, little has been published about systems that can generate\nrecommendations in response to changes in recommendable items and user\nbehaviour in a very short space of time. In this paper we describe a new\nalgorithm for updating collaborative filtering models incrementally, and\ndemonstrate its effectiveness on clickstream data from The Times. We also\ndescribe the architecture that allows recommendations to be generated on the\nfly, and how we have made each component scalable. The system is currently\nbeing used in production at News UK.\n",
"title": "Algorithms and Architecture for Real-time Recommendations at News UK"
}
| null | null |
[
"Computer Science"
] | null | true | null |
12906
| null |
Validated
| null | null |
null |
{
"abstract": " The measurement problem and three other vexing experiments in quantum physics\nare described. It is shown how Quantum Field Theory, as formulated by Julian\nSchwinger, provides simple solutions for all four experiments. It is also shown\nhow this theory resolves many other problems of Quantum Mechanics and\nRelativity, including a new and simple derivation of E = mc2.\n",
"title": "A Physics Tragedy"
}
| null | null | null | null | true | null |
12907
| null |
Default
| null | null |
null |
{
"abstract": " We describe dynamical symmetry breaking in a system of massless Dirac\nfermions with both electromagnetic and four-fermion interactions in (2+1)\ndimensions. The former is described by the Pseudo Quantum Electrodynamics\n(PQED) and the latter is given by the so-called Gross-Neveu action. We apply\nthe Hubbard-Stratonovich transformation and the large$-N_f$ expansion in our\nmodel to obtain a Yukawa action. Thereafter, the presence of a symmetry broken\nphase is inferred from the non-perturbative Schwinger-Dyson equation for the\nelectron propagator. This is the physical solution whenever the fine-structure\nconstant is larger than a critical value $\\alpha_c(D N_f)$. In particular, we\nobtain the critical coupling constant $\\alpha_c\\approx 0.36$ for $D N_f=8$.,\nwhere $D=2,4$ corresponds to the SU(2) and SU(4) cases, respectively, and $N_f$\nis the flavor number. Our results show a decreasing of the critical coupling\nconstant in comparison with the case of pure electromagnetic interaction, thus\nyielding a more favorable scenario for the occurrence of dynamical symmetry\nbreaking. For two-dimensional materials,in application in condensed matter\nsystems, it implies an energy gap at the Dirac points or valleys of the\nhoneycomb lattice.\n",
"title": "Dynamical Mass Generation in Pseudo Quantum Electrodynamics with Four-Fermion Interactions"
}
| null | null | null | null | true | null |
12908
| null |
Default
| null | null |
null |
{
"abstract": " Real world networks are often subject to severe uncertainties which need to\nbe addressed by any reliable prescriptive model. In the context of the maximum\nflow problem subject to arc failure, robust models have gained particular\nattention. For a path-based model, the resulting optimization problem is\nassumed to be difficult in the literature, yet the complexity status is widely\nunknown. We present a computational approach to solve the robust flow problem\nto optimality by simultaneous primal and dual separation, the practical\nefficacy of which is shown by a computational study.\nFurthermore, we introduce a novel model of robust flows which provides a\ncompromise between stochastic and robust optimization by assigning\nprobabilities to groups of scenarios. The new model can be solved by the same\ncomputational techniques as the robust model. A bound on the generalization\nerror is proven for the case that the probabilities are determined empirically.\nThe suggested model as well as the computational approach extend to linear\noptimization problems more general than robust flows.\n",
"title": "Computational Methods for Path-based Robust Flows"
}
| null | null | null | null | true | null |
12909
| null |
Default
| null | null |
null |
{
"abstract": " The use of game theory in the design and control of large scale networked\nsystems is becoming increasingly more important. In this paper, we follow this\napproach to efficiently solve a network allocation problem motivated by\npeer-to- peer cloud storage models as alternatives to classical centralized\ncloud storage services. To this aim, we propose an allocation algorithm that\nallows the units to use their neighbors to store a back up of their data. We\nprove convergence, characterize the final allocation, and corroborate our\nanalysis with extensive numerical simulation that shows the good performance of\nthe algorithm in terms of scalability, complexity and structure of the\nsolution.\n",
"title": "A game theoretic approach to a network cloud storage problem"
}
| null | null | null | null | true | null |
12910
| null |
Default
| null | null |
null |
{
"abstract": " Evolutionary modeling applications are the best way to provide full\ninformation to support in-depth understanding of evaluation of organisms. These\napplications mainly depend on identifying the evolutionary history of existing\norganisms and understanding the relations between them, which is possible\nthrough the deep analysis of their biological sequences. Multiple Sequence\nAlignment (MSA) is considered an important tool in such applications, where it\ngives an accurate representation of the relations between different biological\nsequences. In literature, many efforts have been put into presenting a new MSA\nalgorithm or even improving existing ones. However, little efforts on\noptimizing parallel MSA algorithms have been done. Nowadays, large datasets\nbecome a reality, and big data become a primary challenge in various fields,\nwhich should be also a new milestone for new bioinformatics algorithms. This\nsurvey presents four of the state-of-the-art parallel MSA algorithms, TCoffee,\nMAFFT, MSAProbs, and M2Align. We provide a detailed discussion of each\nalgorithm including its strengths, weaknesses, and implementation details and\nthe effectiveness of its parallel implementation compared to the other\nalgorithms, taking into account the MSA accuracy on two different datasets,\nBAliBASE and OXBench.\n",
"title": "A Survey of the State-of-the-Art Parallel Multiple Sequence Alignment Algorithms on Multicore Systems"
}
| null | null | null | null | true | null |
12911
| null |
Default
| null | null |
null |
{
"abstract": " The Wasserstein distance received a lot of attention recently in the\ncommunity of machine learning, especially for its principled way of comparing\ndistributions. It has found numerous applications in several hard problems,\nsuch as domain adaptation, dimensionality reduction or generative models.\nHowever, its use is still limited by a heavy computational cost. Our goal is to\nalleviate this problem by providing an approximation mechanism that allows to\nbreak its inherent complexity. It relies on the search of an embedding where\nthe Euclidean distance mimics the Wasserstein distance. We show that such an\nembedding can be found with a siamese architecture associated with a decoder\nnetwork that allows to move from the embedding space back to the original input\nspace. Once this embedding has been found, computing optimization problems in\nthe Wasserstein space (e.g. barycenters, principal directions or even\narchetypes) can be conducted extremely fast. Numerical experiments supporting\nthis idea are conducted on image datasets, and show the wide potential benefits\nof our method.\n",
"title": "Learning Wasserstein Embeddings"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
12912
| null |
Validated
| null | null |
null |
{
"abstract": " We provide the first information theoretic tight analysis for inference of\nlatent community structure given a sparse graph along with high dimensional\nnode covariates, correlated with the same latent communities. Our work bridges\nrecent theoretical breakthroughs in the detection of latent community structure\nwithout nodes covariates and a large body of empirical work using diverse\nheuristics for combining node covariates with graphs for inference. The\ntightness of our analysis implies in particular, the information theoretical\nnecessity of combining the different sources of information. Our analysis holds\nfor networks of large degrees as well as for a Gaussian version of the model.\n",
"title": "Contextual Stochastic Block Models"
}
| null | null | null | null | true | null |
12913
| null |
Default
| null | null |
null |
{
"abstract": " Low-rank tensor regression, a new model class that learns high-order\ncorrelation from data, has recently received considerable attention. At the\nsame time, Gaussian processes (GP) are well-studied machine learning models for\nstructure learning. In this paper, we demonstrate interesting connections\nbetween the two, especially for multi-way data analysis. We show that low-rank\ntensor regression is essentially learning a multi-linear kernel in Gaussian\nprocesses, and the low-rank assumption translates to the constrained Bayesian\ninference problem. We prove the oracle inequality and derive the average case\nlearning curve for the equivalent GP model. Our finding implies that low-rank\ntensor regression, though empirically successful, is highly dependent on the\neigenvalues of covariance functions as well as variable correlations.\n",
"title": "Tensor Regression Meets Gaussian Processes"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
12914
| null |
Validated
| null | null |
null |
{
"abstract": " The Canonical Polyadic decomposition (CPD) is a convenient and intuitive tool\nfor tensor factorization; however, for higher-order tensors, it often exhibits\nhigh computational cost and permutation of tensor entries, these undesirable\neffects grow exponentially with the tensor order. Prior compression of tensor\nin-hand can reduce the computational cost of CPD, but this is only applicable\nwhen the rank $R$ of the decomposition does not exceed the tensor dimensions.\nTo resolve these issues, we present a novel method for CPD of higher-order\ntensors, which rests upon a simple tensor network of representative\ninter-connected core tensors of orders not higher than 3. For rigour, we\ndevelop an exact conversion scheme from the core tensors to the factor matrices\nin CPD, and an iterative algorithm with low complexity to estimate these factor\nmatrices for the inexact case. Comprehensive simulations over a variety of\nscenarios support the approach.\n",
"title": "Tensor Networks for Latent Variable Analysis: Higher Order Canonical Polyadic Decomposition"
}
| null | null | null | null | true | null |
12915
| null |
Default
| null | null |
null |
{
"abstract": " In this study, the authors develop a structural model that combines a macro\ndiffusion model with a micro choice model to control for the effect of social\ninfluence on the mobile app choices of customers over app stores. Social\ninfluence refers to the density of adopters within the proximity of other\ncustomers. Using a large data set from an African app store and Bayesian\nestimation methods, the authors quantify the effect of social influence and\ninvestigate the impact of ignoring this process in estimating customer choices.\nThe findings show that customer choices in the app store are explained better\nby offline than online density of adopters and that ignoring social influence\nin estimations results in biased estimates. Furthermore, the findings show that\nthe mobile app adoption process is similar to adoption of music CDs, among all\nother classic economy goods. A counterfactual analysis shows that the app store\ncan increase its revenue by 13.6% through a viral marketing policy (e.g., a\nsharing with friends and family button).\n",
"title": "Social Learning and Diffusion of Pervasive Goods: An Empirical Study of an African App Store"
}
| null | null | null | null | true | null |
12916
| null |
Default
| null | null |
null |
{
"abstract": " The Empirical Mode Decomposition (EMD) provides a tool to characterize time\nseries in terms of its implicit components oscillating at different\ntime-scales. We apply this decomposition to intraday time series of the\nfollowing three financial indices: the S\\&P 500 (USA), the IPC (Mexico) and the\nVIX (volatility index USA), obtaining time-varying multidimensional\ncross-correlations at different time-scales. The correlations computed over a\nrolling window are compared across the three indices, across the components at\ndifferent time-scales, at different lags and over time. We uncover a rich\nheterogeneity of interactions which depends on the time-scale and has important\nled-lag relations which can have practical use for portfolio management, risk\nestimation and investments.\n",
"title": "Dynamic correlations at different time-scales with Empirical Mode Decomposition"
}
| null | null |
[
"Computer Science"
] | null | true | null |
12917
| null |
Validated
| null | null |
null |
{
"abstract": " We study equilibrium properties of catalytically-activated $A + A \\to\n\\oslash$ reactions taking place on a lattice of adsorption sites. The particles\nundergo continuous exchanges with a reservoir maintained at a constant chemical\npotential $\\mu$ and react when they appear at the neighbouring sites, provided\nthat some reactive conditions are fulfilled. We model the latter in two\ndifferent ways: In the Model I some fraction $p$ of the {\\em bonds} connecting\nneighbouring sites possesses special catalytic properties such that any two\n$A$s appearing on the sites connected by such a bond instantaneously react and\ndesorb. In the Model II some fraction $p$ of the adsorption {\\em sites}\npossesses such properties and neighbouring particles react if at least one of\nthem resides on a catalytic site. For the case of \\textit{annealed} disorder in\nthe distribution of the catalyst, which is tantamount to the situation when the\nreaction may take place at any point on the lattice but happens with a finite\nprobability $p$, we provide an exact solution for both models for the interior\nof an infinitely large Cayley tree - the so-called Bethe lattice. We show that\nboth models exhibit a rich critical behaviour: For the annealed Model I it is\ncharacterised by a transition into an ordered state and a re-entrant transition\ninto a disordered phase, which both are continuous. For the annealed Model II,\nwhich represents a rather exotic model of statistical mechanics in which\ninteractions of any particle with its environment have a peculiar Boolean form,\nthe transition to an ordered state is always continuous, while the re-entrant\ntransition into the disordered phase may be either continuous or discontinuous,\ndepending on the value of $p$.\n",
"title": "Order-disorder transitions in lattice gases with annealed reactive constraints"
}
| null | null | null | null | true | null |
12918
| null |
Default
| null | null |
null |
{
"abstract": " Convolutional Neural Networks (CNNs) have become the method of choice for\nlearning problems involving 2D planar images. However, a number of problems of\nrecent interest have created a demand for models that can analyze spherical\nimages. Examples include omnidirectional vision for drones, robots, and\nautonomous cars, molecular regression problems, and global weather and climate\nmodelling. A naive application of convolutional networks to a planar projection\nof the spherical signal is destined to fail, because the space-varying\ndistortions introduced by such a projection will make translational weight\nsharing ineffective.\nIn this paper we introduce the building blocks for constructing spherical\nCNNs. We propose a definition for the spherical cross-correlation that is both\nexpressive and rotation-equivariant. The spherical correlation satisfies a\ngeneralized Fourier theorem, which allows us to compute it efficiently using a\ngeneralized (non-commutative) Fast Fourier Transform (FFT) algorithm. We\ndemonstrate the computational efficiency, numerical accuracy, and effectiveness\nof spherical CNNs applied to 3D model recognition and atomization energy\nregression.\n",
"title": "Spherical CNNs"
}
| null | null | null | null | true | null |
12919
| null |
Default
| null | null |
null |
{
"abstract": " The recent observations of rippled structures on the surface of the Orion\nmolecular cloud (Berné et al. 2010), have been attributed to the\nKelvin-Helmholtz (KH) instability. The wavelike structures which have mainly\nseen near star-forming regions taking place at the interface between the hot\ndiffuse gas, which is ionized by massive stars, and the cold dense molecular\nclouds. The radiation pressure of massive stars and stellar clusters is one of\nthe important issues that has been considered frequently in the dynamics of\nclouds. Here, we investigate the influence of radiation pressure, from\nwell-known Trapezium cluster in the Orion nebula, on the evolution of KH\ninstability. The stability of the interface between HII region and molecular\nclouds in the presence of the radiation pressure, has been studied using the\nlinear perturbation analysis for the certain range of the wavelengths. The\nlinear analysis show that consideration of the radiation pressure intensifies\nthe growth rate of KH modes and consequently decreases the e-fold time-scale of\nthe instability. On the other hand the domain of the instability is extended\nand includes the more wavelengths, consisting of smaller ones rather than the\ncase when the effect of the radiation pressure is not considered. Our results\nshows that for $\\lambda_{\\rm KH}>0.15\\rm pc$, the growth rate of KH instability\ndose not depend to the radiation pressure. Based on our results, the radiation\npressure is a triggering mechanism in development of the KH instability and\nsubsequently formation of turbulent sub-structures in the molecular clouds near\nmassive stars. The role of magnetic fields in the presence of the radiation\npressure is also investigated and it is resulted that the magnetic field\nsuppresses the effects induced by the radiation pressure.\n",
"title": "The Kelvin-Helmholtz instability in the Orion nebula: The effect of radiation pressure"
}
| null | null |
[
"Physics"
] | null | true | null |
12920
| null |
Validated
| null | null |
null |
{
"abstract": " Si Li and author suggested in that, in some cases, the AdS/CFT correspondence\ncan be formulated in terms of the algebraic operation of Koszul duality. In\nthis paper this suggestion is checked explicitly for $M2$ branes in an\n$\\Omega$-background. The algebra of supersymmetric operators on a stack of $K$\n$M2$ branes is shown to be Koszul dual, in large $K$, to the algebra of\nsupersymmetric operators of $11$-dimensional supergravity in an\n$\\Omega$-background (using the formulation of supergravity in an\n$\\Omega$-background presented in arXiv:1610.04144).\nThe twisted form of supergravity that is used here can be quantized to all\norders in perturbation theory. We find that the Koszul duality result holds to\nall orders in perturbation theory, in both the gravitational theory and the\ntheory on the $M2$. (However, there is a certain non-linear identification of\nthe coupling constants on each side which I was unable to determine\nexplicitly).\nIt is also shown that the algebra of operators on $K$ $M2$ branes, as $K \\to\n\\infty$, is a quantum double-loop algebra (a two-variable analog of the\nYangian). This algebra is also the Koszul dual of the algebra of operators on\nthe gravitational theory. An explicit presentation for this algebra is\npresented, and it is shown that this algebra is the unique quantization of its\nclassical limit. Some conjectural applications to enumerative geometry of\nCalabi-Yau threefolds are also presented.\n",
"title": "Holography and Koszul duality: the example of the $M2$ brane"
}
| null | null | null | null | true | null |
12921
| null |
Default
| null | null |
null |
{
"abstract": " This paper is devoted to the uniqueness problem of the power of a meromorphic\nfunction with its differential polynomial sharing a set. Our result will extend\na number of results obtained in the theory of normal families. Some questions\nare posed for future research.\n",
"title": "Uniqueness of the power of a meromorphic functions with its differential polynomial sharing a set"
}
| null | null | null | null | true | null |
12922
| null |
Default
| null | null |
null |
{
"abstract": " We measure the gate voltage ($V_g$) dependence of the superconducting\nproperties and the spin-orbit interaction in the (111)-oriented\nLaAlO$_3$/SrTiO$_3$ interface. Superconductivity is observed in a dome-shaped\nregion in the carrier density-temperature phase diagram with the maxima of\nsuperconducting transition temperature $T_c$ and the upper critical fields\nlying at the same $V_g$. The spin-orbit interaction determined from the\nsuperconducting parameters and confirmed by weak-antilocalization measurements\nfollows the same gate voltage dependence as $T_c$. The correlation between the\nsuperconductivity and spin-orbit interaction as well as the enhancement of the\nparallel upper critical field, well beyond the Chandrasekhar-Clogston limit\nsuggest that superconductivity and the spin-orbit interaction are linked in a\nnontrivial fashion. We propose possible scenarios to explain this\nunconventional behavior.\n",
"title": "Link between the Superconducting Dome and Spin-Orbit Interaction in the (111) LaAlO$_3$/SrTiO$_3$ Interface"
}
| null | null |
[
"Physics"
] | null | true | null |
12923
| null |
Validated
| null | null |
null |
{
"abstract": " In this work we review a class of deterministic nonlinear models for the\npropagation of infectious diseases over contact networks with\nstrongly-connected topologies. We consider network models for\nsusceptible-infected (SI), susceptible-infected-susceptible (SIS), and\nsusceptible-infected-recovered (SIR) settings. In each setting, we provide a\ncomprehensive nonlinear analysis of equilibria, stability properties,\nconvergence, monotonicity, positivity, and threshold conditions. For the\nnetwork SI setting, specific contributions include establishing its equilibria,\nstability, and positivity properties. For the network SIS setting, we review a\nwell-known deterministic model, provide novel results on the computation and\ncharacterization of the endemic state (when the system is above the epidemic\nthreshold), and present alternative proofs for some of its properties. Finally,\nfor the network SIR setting, we propose novel results for transient behavior,\nthreshold conditions, stability properties, and asymptotic convergence. These\nresults are analogous to those well-known for the scalar case. In addition, we\nprovide a novel iterative algorithm to compute the asymptotic state of the\nnetwork SIR system.\n",
"title": "On the Dynamics of Deterministic Epidemic Propagation over Networks"
}
| null | null | null | null | true | null |
12924
| null |
Default
| null | null |
null |
{
"abstract": " Text-dependent speaker verification is becoming popular in the speaker\nrecognition society. However, the conventional i-vector framework which has\nbeen successful for speaker identification and other similar tasks works\nrelatively poorly in this task. Researchers have proposed several new methods\nto improve performance, but it is still unclear that which model is the best\nchoice, especially when the pass-phrases are prompted during enrollment and\ntest. In this paper, we introduce four modeling methods and compare their\nperformance on the newly published RedDots dataset. To further explore the\ninfluence of different frame alignments, Viterbi and forward-backward\nalgorithms are both used in the HMM-based models. Several bottleneck features\nare also investigated. Our experiments show that, by explicitly modeling the\nlexical content, the HMM-based modeling achieves good results in the\nfixed-phrase condition. In the prompted-phrase condition, GMM-HMM and\ni-vector/HMM are not as successful. In both conditions, the forward-backward\nalgorithm brings more benefits to the i-vector/HMM system. Additionally, we\nalso find that even though bottleneck features perform well for\ntext-independent speaker verification, they do not outperform MFCCs on the most\nchallenging Imposter-Correct trials on RedDots.\n",
"title": "Comparison of Multiple Features and Modeling Methods for Text-dependent Speaker Verification"
}
| null | null | null | null | true | null |
12925
| null |
Default
| null | null |
null |
{
"abstract": " Lactate threshold is considered an essential parameter when assessing\nperformance of elite and recreational runners and prescribing training\nintensities in endurance sports. However, the measurement of blood lactate\nconcentration requires expensive equipment and the extraction of blood samples,\nwhich are inconvenient for frequent monitoring. Furthermore, most recreational\nrunners do not have access to routine assessment of their physical fitness by\nthe aforementioned equipment so they are not able to calculate the lactate\nthreshold without resorting to an expensive and specialized centre. Therefore,\nthe main objective of this study is to create an intelligent system capable of\nestimating the lactate threshold of recreational athletes participating in\nendurance running sports. The solution here proposed is based on a machine\nlearning system which models the lactate evolution using recurrent neural\nnetworks and includes the proposal of standardization of the temporal axis as\nwell as a modification of the stratified sampling method. The results show that\nthe proposed system accurately estimates the lactate threshold of 89.52% of the\nathletes and its correlation with the experimentally measured lactate threshold\nis very high (R=0,89). Moreover, its behaviour with the test dataset is as good\nas with the training set, meaning that the generalization power of the model is\nhigh. Therefore, in this study a machine learning based system is proposed as\nalternative to the traditional invasive lactate threshold measurement tests for\nrecreational runners.\n",
"title": "Estimation of lactate threshold with machine learning techniques in recreational runners"
}
| null | null | null | null | true | null |
12926
| null |
Default
| null | null |
null |
{
"abstract": " It is shown that an equiprobability hypothesis leads to a scenario in which\nit is possible to predict the outcome of a single toss of a fair coin with a\nsuccess probability greater than 50%. We discuss whether this hypothesis might\nbe independent of the usual hypotheses governing probability, as well as\nwhether this hypothesis might be assumed as a result of the Principle of\nIndifference. Also discussed are ways to implement or circumvent the\nhypothesis.\n",
"title": "A Coin-Tossing Conundrum"
}
| null | null |
[
"Statistics"
] | null | true | null |
12927
| null |
Validated
| null | null |
null |
{
"abstract": " We construct a special class of Lorentz surfaces in the pseudo-Euclidean\n4-space with neutral metric which are one-parameter systems of meridians of\nrotational hypersurfaces with lightlike axis and call them meridian surfaces.\nWe give the complete classification of the meridian surfaces with constant\nGauss curvature and prove that there are no meridian surfaces with parallel\nmean curvature vector field other than CMC surfaces lying in a hyperplane. We\nalso classify the meridian surfaces with parallel normalized mean curvature\nvector field. We show that in the family of the meridian surfaces there exist\nLorentz surfaces which have parallel normalized mean curvature vector field but\nnot parallel mean curvature vector.\n",
"title": "Meridian Surfaces on Rotational Hypersurfaces with Lightlike Axis in ${\\mathbb E}^4_2$"
}
| null | null | null | null | true | null |
12928
| null |
Default
| null | null |
null |
{
"abstract": " Since the limited power capacity, finite inertia, and dynamic loads make the\nshipboard power system (SPS) vulnerable, the automatic reconfiguration for\nfailure recovery in SPS is an extremely significant but still challenging\nproblem. It is not only required to operate accurately and optimally, but also\nto satisfy operating constraints. In this paper, we consider the\nreconfiguration optimization for hybrid AC/DC microgrids in all-electric ships.\nFirstly, the multi-zone medium voltage DC (MVDC) SPS model is presented. In\nthis model, the DC power flow for reconfiguration and a generalized AC/DC\nconverter are modeled for accurate reconfiguration. Secondly, since this\nproblem is mixed integer nonlinear programming (MINLP), a hybrid method based\non Newton Raphson and Biogeography based Optimization (NRBBO) is designed\naccording to the characteristics of system, loads, and faults. This method\nfacilitates to maximize the weighted load restoration while satisfying\noperating constraints. Finally, the simulation results demonstrate this method\nhas advantages in terms of power restoration and convergence speed.\n",
"title": "Hybrid Optimization Method for Reconfiguration of AC/DC Microgrids in All-Electric Ships"
}
| null | null | null | null | true | null |
12929
| null |
Default
| null | null |
null |
{
"abstract": " This paper extends fully-convolutional neural networks (FCN) for the clothing\nparsing problem. Clothing parsing requires higher-level knowledge on clothing\nsemantics and contextual cues to disambiguate fine-grained categories. We\nextend FCN architecture with a side-branch network which we refer outfit\nencoder to predict a consistent set of clothing labels to encourage\ncombinatorial preference, and with conditional random field (CRF) to explicitly\nconsider coherent label assignment to the given image. The empirical results\nusing Fashionista and CFPD datasets show that our model achieves\nstate-of-the-art performance in clothing parsing, without additional\nsupervision during training. We also study the qualitative influence of\nannotation on the current clothing parsing benchmarks, with our Web-based tool\nfor multi-scale pixel-wise annotation and manual refinement effort to the\nFashionista dataset. Finally, we show that the image representation of the\noutfit encoder is useful for dress-up image retrieval application.\n",
"title": "Looking at Outfit to Parse Clothing"
}
| null | null |
[
"Computer Science"
] | null | true | null |
12930
| null |
Validated
| null | null |
null |
{
"abstract": " Ba$_8$CoNb$_6$O$_{24}$ presents a system whose Co$^{2+}$ ions have an\neffective spin 1/2 and construct a regular triangular-lattice antiferromagnet\n(TLAFM) with a very large interlayer spacing, ensuring purely two-dimensional\ncharacter. We exploit this ideal realization to perform a detailed experimental\nanalysis of the $S = 1/2$ TLAFM, which is one of the keystone models in\nfrustrated quantum magnetism. We find strong low-energy spin fluctuations and\nno magnetic ordering, but a diverging correlation length down to 0.1 K,\nindicating a Mermin-Wagner trend towards zero-temperature order. Below 0.1 K,\nhowever, our low-field measurements show an nexpected magnetically disordered\nstate, which is a candidate quantum spin liquid. We establish the $(H,T)$ phase\ndiagram, mapping in detail the quantum fluctuation corrections to the available\ntheoretical analysis. These include a strong upshift in field of the maximum\nordering temperature, qualitative changes to both low- and high-field phase\nboundaries, and an ordered regime apparently dominated by the collinear\n\"up-up-down\" state. Ba$_8$CoNb$_6$O$_{24}$ therefore offers fresh input for the\ndevelopment of theoretical approaches to the field-induced quantum phase\ntransitions of the $S = 1/2$ Heisenberg TLAFM.\n",
"title": "Mermin-Wagner physics, (H,T) phase diagram, and candidate quantum spin-liquid phase in the spin-1/2 triangular-lattice antiferromagnet Ba8CoNb6O24"
}
| null | null |
[
"Physics"
] | null | true | null |
12931
| null |
Validated
| null | null |
null |
{
"abstract": " A regular language $L$ is non-returning if in the minimal deterministic\nfinite automaton accepting it there are no transitions into the initial state.\nEom, Han and Jirásková derived upper bounds on the state complexity of\nboolean operations and Kleene star, and proved that these bounds are tight\nusing two different binary witnesses. They derived upper bounds for\nconcatenation and reversal using three different ternary witnesses. These five\nwitnesses use a total of six different transformations. We show that for each\n$n\\ge 4$ there exists a ternary witness of state complexity $n$ that meets the\nbound for reversal and that at least three letters are needed to meet this\nbound. Moreover, the restrictions of this witness to binary alphabets meet the\nbounds for product, star, and boolean operations. We also derive tight upper\nbounds on the state complexity of binary operations that take arguments with\ndifferent alphabets. We prove that the maximal syntactic semigroup of a\nnon-returning language has $(n-1)^n$ elements and requires at least\n$\\binom{n}{2}$ generators. We find the maximal state complexities of atoms of\nnon-returning languages. Finally, we show that there exists a most complex\nnon-returning language that meets the bounds for all these complexity measures.\n",
"title": "Most Complex Non-Returning Regular Languages"
}
| null | null | null | null | true | null |
12932
| null |
Default
| null | null |
null |
{
"abstract": " This paper proposes an extension to the Generative Adversarial Networks\n(GANs), namely as ARTGAN to synthetically generate more challenging and complex\nimages such as artwork that have abstract characteristics. This is in contrast\nto most of the current solutions that focused on generating natural images such\nas room interiors, birds, flowers and faces. The key innovation of our work is\nto allow back-propagation of the loss function w.r.t. the labels (randomly\nassigned to each generated images) to the generator from the discriminator.\nWith the feedback from the label information, the generator is able to learn\nfaster and achieve better generated image quality. Empirically, we show that\nthe proposed ARTGAN is capable to create realistic artwork, as well as generate\ncompelling real world images that globally look natural with clear shape on\nCIFAR-10.\n",
"title": "ArtGAN: Artwork Synthesis with Conditional Categorical GANs"
}
| null | null | null | null | true | null |
12933
| null |
Default
| null | null |
null |
{
"abstract": " Working over an infinite field of positive characteristic, an upper bound is\ngiven for the nilpotency index of a finitely generated nil algebra of bounded\nnil index $n$ in terms of the maximal degree in a minimal homogenous generating\nsystem of the ring of simultaneous conjugation invariants of tuples of $n$ by\n$n$ matrices. This is deduced from a result of Zubkov. As a consequence, a\nrecent degree bound due to Derksen and Makam for the generators of the ring of\nmatrix invariants yields an upper bound for the nilpotency index of a finitely\ngenerated nil algebra that is polynomial in the number of generators and the\nnil index. Furthermore, a characteristic free treatment is given to Kuzmin's\nlower bound for the nilpotency index.\n",
"title": "Polynomial bound for the nilpotency index of finitely generated nil algebras"
}
| null | null |
[
"Mathematics"
] | null | true | null |
12934
| null |
Validated
| null | null |
null |
{
"abstract": " We present CFAAR, a program repair assistance technique that operates by\nselectively altering the outcome of suspicious predicates in order to yield\nexpected behavior. CFAAR is applicable to defects that are repairable by\nnegating predicates under specific conditions. CFAAR proceeds as follows: 1) it\nidentifies predicates such that negating them at given instances would make the\nfailing tests exhibit correct behavior; 2) for each candidate predicate, it\nuses the program state information to build a classifier that dictates when the\npredicate should be negated; 3) for each classifier, it leverages a Decision\nTree to synthesize a patch to be presented to the developer. We evaluated our\ntoolset using 149 defects from the IntroClass and Siemens benchmarks. CFAAR\nidentified 91 potential candidate defects and generated plausible patches for\n41 of them. Twelve of the patches are believed to be correct, whereas the rest\nprovide repair assistance to the developer.\n",
"title": "CFAAR: Control Flow Alteration to Assist Repair"
}
| null | null | null | null | true | null |
12935
| null |
Default
| null | null |
null |
{
"abstract": " Past work in relation extraction has focused on binary relations in single\nsentences. Recent NLP inroads in high-value domains have sparked interest in\nthe more general setting of extracting n-ary relations that span multiple\nsentences. In this paper, we explore a general relation extraction framework\nbased on graph long short-term memory networks (graph LSTMs) that can be easily\nextended to cross-sentence n-ary relation extraction. The graph formulation\nprovides a unified way of exploring different LSTM approaches and incorporating\nvarious intra-sentential and inter-sentential dependencies, such as sequential,\nsyntactic, and discourse relations. A robust contextual representation is\nlearned for the entities, which serves as input to the relation classifier.\nThis simplifies handling of relations with arbitrary arity, and enables\nmulti-task learning with related relations. We evaluate this framework in two\nimportant precision medicine settings, demonstrating its effectiveness with\nboth conventional supervised learning and distant supervision. Cross-sentence\nextraction produced larger knowledge bases. and multi-task learning\nsignificantly improved extraction accuracy. A thorough analysis of various LSTM\napproaches yielded useful insight the impact of linguistic analysis on\nextraction accuracy.\n",
"title": "Cross-Sentence N-ary Relation Extraction with Graph LSTMs"
}
| null | null | null | null | true | null |
12936
| null |
Default
| null | null |
null |
{
"abstract": " Users organize themselves into communities on web platforms. These\ncommunities can interact with one another, often leading to conflicts and toxic\ninteractions. However, little is known about the mechanisms of interactions\nbetween communities and how they impact users.\nHere we study intercommunity interactions across 36,000 communities on\nReddit, examining cases where users of one community are mobilized by negative\nsentiment to comment in another community. We show that such conflicts tend to\nbe initiated by a handful of communities---less than 1% of communities start\n74% of conflicts. While conflicts tend to be initiated by highly active\ncommunity members, they are carried out by significantly less active members.\nWe find that conflicts are marked by formation of echo chambers, where users\nprimarily talk to other users from their own community. In the long-term,\nconflicts have adverse effects and reduce the overall activity of users in the\ntargeted communities.\nOur analysis of user interactions also suggests strategies for mitigating the\nnegative impact of conflicts---such as increasing direct engagement between\nattackers and defenders. Further, we accurately predict whether a conflict will\noccur by creating a novel LSTM model that combines graph embeddings, user,\ncommunity, and text features. This model can be used toreate early-warning\nsystems for community moderators to prevent conflicts. Altogether, this work\npresents a data-driven view of community interactions and conflict, and paves\nthe way towards healthier online communities.\n",
"title": "Community Interaction and Conflict on the Web"
}
| null | null | null | null | true | null |
12937
| null |
Default
| null | null |
null |
{
"abstract": " Recurrence networks and the associated statistical measures have become\nimportant tools in the analysis of time series data. In this work, we test how\neffective the recurrence network measures are in analyzing real world data\ninvolving two main types of noise, white noise and colored noise. We use two\nprominent network measures as discriminating statistic for hypothesis testing\nusing surrogate data for a specific null hypothesis that the data is derived\nfrom a linear stochastic process. We show that the characteristic path length\nis especially efficient as a discriminating measure with the conclusions\nreasonably accurate even with limited number of data points in the time series.\nWe also highlight an additional advantage of the network approach in\nidentifying the dimensionality of the system underlying the time series through\na convergence measure derived from the probability distribution of the local\nclustering coefficients. As examples of real world data, we use the light\ncurves from a prominent black hole system and show that a combined analysis\nusing three primary network measures can provide vital information regarding\nthe nature of temporal variability of light curves from different spectroscopic\nclasses.\n",
"title": "Recurrence network measures for hypothesis testing using surrogate data: application to black hole light curves"
}
| null | null |
[
"Physics"
] | null | true | null |
12938
| null |
Validated
| null | null |
null |
{
"abstract": " Traditional optical imaging faces an unavoidable trade-off between resolution\nand depth of field (DOF). To increase resolution, high numerical apertures (NA)\nare needed, but the associated large angular uncertainty results in a limited\nrange of depths that can be put in sharp focus. Plenoptic imaging was\nintroduced a few years ago to remedy this trade off. To this aim, plenoptic\nimaging reconstructs the path of light rays from the lens to the sensor.\nHowever, the improvement offered by standard plenoptic imaging is practical and\nnot fundamental: the increased DOF leads to a proportional reduction of the\nresolution well above the diffraction limit imposed by the lens NA. In this\npaper, we demonstrate that correlation measurements enable pushing plenoptic\nimaging to its fundamental limits of both resolution and DOF. Namely, we\ndemonstrate to maintain the imaging resolution at the diffraction limit while\nincreasing the depth of field by a factor of 7. Our results represent the\ntheoretical and experimental basis for the effective development of the\npromising applications of plenoptic imaging.\n",
"title": "Diffraction-limited plenoptic imaging with correlated light"
}
| null | null | null | null | true | null |
12939
| null |
Default
| null | null |
null |
{
"abstract": " Complex performance measures, beyond the popular measure of accuracy, are\nincreasingly being used in the context of binary classification. These complex\nperformance measures are typically not even decomposable, that is, the loss\nevaluated on a batch of samples cannot typically be expressed as a sum or\naverage of losses evaluated at individual samples, which in turn requires new\ntheoretical and methodological developments beyond standard treatments of\nsupervised learning. In this paper, we advance this understanding of binary\nclassification for complex performance measures by identifying two key\nproperties: a so-called Karmic property, and a more technical\nthreshold-quasi-concavity property, which we show is milder than existing\nstructural assumptions imposed on performance measures. Under these properties,\nwe show that the Bayes optimal classifier is a threshold function of the\nconditional probability of positive class. We then leverage this result to come\nup with a computationally practical plug-in classifier, via a novel threshold\nestimator, and further, provide a novel statistical analysis of classification\nerror with respect to complex performance measures.\n",
"title": "Binary Classification with Karmic, Threshold-Quasi-Concave Metrics"
}
| null | null | null | null | true | null |
12940
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we consider a novel machine learning problem, that is,\nlearning a classifier from noisy label distributions. In this problem, each\ninstance with a feature vector belongs to at least one group. Then, instead of\nthe true label of each instance, we observe the label distribution of the\ninstances associated with a group, where the label distribution is distorted by\nan unknown noise. Our goals are to (1) estimate the true label of each\ninstance, and (2) learn a classifier that predicts the true label of a new\ninstance. We propose a probabilistic model that considers true label\ndistributions of groups and parameters that represent the noise as hidden\nvariables. The model can be learned based on a variational Bayesian method. In\nnumerical experiments, we show that the proposed model outperforms existing\nmethods in terms of the estimation of the true labels of instances.\n",
"title": "Learning from Noisy Label Distributions"
}
| null | null | null | null | true | null |
12941
| null |
Default
| null | null |
null |
{
"abstract": " Phased Array Feed (PAF) technology is the next major advancement in radio\nastronomy in terms of combining high sensitivity and large field of view. The\nFocal L-band Array for the Green Bank Telescope (FLAG) is one of the most\nsensitive PAFs developed so far. It consists of 19 dual-polarization elements\nmounted on a prime focus dewar resulting in seven beams on the sky. Its\nunprecedented system temperature of$\\sim$17 K will lead to a 3 fold increase in\npulsar survey speeds as compared to contemporary single pixel feeds. Early\nscience observations were conducted in a recently concluded commissioning phase\nof the FLAG where we clearly demonstrated its science capabilities. We observed\na selection of normal and millisecond pulsars and detected giant pulses from\nPSR B1937+21.\n",
"title": "Commissioning of FLAG: A phased array feed for the GBT"
}
| null | null | null | null | true | null |
12942
| null |
Default
| null | null |
null |
{
"abstract": " Let $X$ be a normal algebraic variety over a finitely generated field $k$ of\ncharacteristic zero, and let $\\ell$ be a prime. Say that a continuous\n$\\ell$-adic representation $\\rho$ of $\\pi_1^{\\text{ét}}(X_{\\bar k})$ is\narithmetic if there exists a representation $\\tilde \\rho$ of a finite index\nsubgroup of $\\pi_1^{\\text{ét}}(X)$, with $\\rho$ a subquotient of\n$\\tilde\\rho|_{\\pi_1(X_{\\bar k})}$. We show that there exists an integer $N=N(X,\n\\ell)$ such that every nontrivial, semisimple arithmetic representation of\n$\\pi_1^{\\text{ét}}(X_{\\bar k})$ is nontrivial mod $\\ell^N$. As a corollary,\nwe prove that any nontrivial semisimple representation of\n$\\pi_1^{\\text{ét}}(X_{\\bar k})$, which arises from geometry, is nontrivial\nmod $\\ell^N$.\n",
"title": "Arithmetic representations of fundamental groups I"
}
| null | null |
[
"Mathematics"
] | null | true | null |
12943
| null |
Validated
| null | null |
null |
{
"abstract": " We introduce and study the problem of optimizing arbitrary functions over\ndegree sequences of hypergraphs and multihypergraphs. We show that over\nmultihypergraphs the problem can be solved in polynomial time. For hypergraphs,\nwe show that deciding if a given sequence is the degree sequence of a\n3-hypergraph is NP-complete, thereby solving a 30 year long open problem. This\nimplies that optimization over hypergraphs is hard already for simple concave\nfunctions. In contrast, we show that for graphs, if the functions at vertices\nare the same, then the problem is polynomial time solvable. We also provide\npositive results for convex optimization over multihypergraphs and graphs and\nexploit connections to degree sequence polytopes and threshold graphs. We then\nelaborate on connections to the emerging theory of shifted combinatorial\noptimization.\n",
"title": "Optimization over Degree Sequences"
}
| null | null |
[
"Computer Science",
"Mathematics"
] | null | true | null |
12944
| null |
Validated
| null | null |
null |
{
"abstract": " We present the results of very long baseline interferometry (VLBI)\nobservations of gamma-ray bright blazar S5 0716+714 using the Korean VLBI\nNetwork (KVN) at the 22, 43, 86, and 129 GHz bands, as part of the\nInterferometric Monitoring of Gamma-ray Bright AGNs (iMOGABA) KVN key science\nprogram. Observations were conducted in 29 sessions from January 16, 2013 to\nMarch 1, 2016, with the source being detected and imaged at all available\nfrequencies. In all epochs, the source was compact on the milliarcsecond (mas)\nscale, yielding a compact VLBI core dominating the synchrotron emission on\nthese scales. Based on the multi-wavelength data between 15 GHz (Owens Valley\nRadio Observatory) and 230 GHz (Submillimeter Array), we found that the source\nshows multiple prominent enhancements of the flux density at the centimeter\n(cm) and millimeter (mm) wavelengths, with mm enhancements leading cm\nenhancements by -16$\\pm$8 days. The turnover frequency was found to vary\nbetween 21 to 69GHz during our observations. By assuming a synchrotron\nself-absorption model for the relativistic jet emission in S5 0716+714, we\nfound the magnetic field strength in the mas emission region to be $\\le$5 mG\nduring the observing period, yielding a weighted mean of 1.0$\\pm$0.6 mG for\nhigher turnover frequencies (e.g., >45 GHz).\n",
"title": "Interferometric Monitoring of Gamma-ray Bright AGNs: S5 0716+714"
}
| null | null | null | null | true | null |
12945
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we study several aspects related with solutions of nonlocal\nproblems whose prototype is $$ u_t =\\displaystyle \\int_{\\mathbb{R}^N} J(x-y)\n\\big( u(y,t) -u(x,t) \\big) \\mathcal G\\big( u(y,t) -u(x,t) \\big) dy \\qquad\n\\mbox{ in } \\, \\Omega \\times (0,T)\\,, $$ being $ u (x,t)=0 \\mbox{ in }\n(\\mathbb{R}^N\\setminus \\Omega )\\times (0,T)\\,$ and $ u(x,0)=u_0 (x) \\mbox{ in }\n\\Omega$. We take, as the most important instance, $\\mathcal G (s) \\sim 1+\n\\frac{\\mu}{2} \\frac{s}{1+\\mu^2 s^2 }$ with $\\mu\\in \\mathbb{R}$ as well as $u_0\n\\in L^1 (\\Omega)$, $J$ is a smooth symmetric function with compact support and\n$\\Omega$ is either a bounded smooth subset of $\\mathbb{R}^N$, with nonlocal\nDirichlet boundary condition, or $\\mathbb{R}^N$ itself.\nThe results deal with existence, uniqueness, comparison principle and\nasymptotic behavior. Moreover we prove that if the kernel rescales in a\nsuitable way, the unique solution of the above problem converges to a solution\nof the deterministic Kardar-Parisi-Zhang equation.\n",
"title": "Parabolic equations with natural growth approximated by nonlocal equations"
}
| null | null | null | null | true | null |
12946
| null |
Default
| null | null |
null |
{
"abstract": " In this work, we analyze the excitonic gap generation in the strong-coupling\nregime of thin films of three-dimensional time-reversal-invariant topological\ninsulators. We start by writing down the effective gauge theory in\n2+1-dimensions from the projection of the 3+1-dimensional quantum\nelectrodynamics. Within this method, we obtain a short-range interaction, which\nhas the form of a Thirring-like term, and a long-range one. The interaction\nbetween the two surface states of the material induces an excitonic gap. By\nusing the large-$N$ approximation in the strong-coupling limit, we find that\nthere is a dynamical mass generation for the excitonic states that preserves\ntime-reversal symmetry and is related to the dynamical chiral-symmetry breaking\nof our model. This symmetry breaking occurs only for values of the\nfermion-flavor number smaller than $N_{c}\\approx 11.8$. Our results show that\nthe inclusion of the full dynamical interaction strongly modifies the critical\nnumber of flavors for the occurrence of exciton condensation, and therefore,\ncannot be neglected.\n",
"title": "Excitonic gap generation in thin-film topological insulators"
}
| null | null | null | null | true | null |
12947
| null |
Default
| null | null |
null |
{
"abstract": " Preference elicitation is the task of suggesting a highly preferred\nconfiguration to a decision maker. The preferences are typically learned by\nquerying the user for choice feedback over pairs or sets of objects. In its\nconstructive variant, new objects are synthesized \"from scratch\" by maximizing\nan estimate of the user utility over a combinatorial (possibly infinite) space\nof candidates. In the constructive setting, most existing elicitation\ntechniques fail because they rely on exhaustive enumeration of the candidates.\nA previous solution explicitly designed for constructive tasks comes with no\nformal performance guarantees, and can be very expensive in (or unapplicable\nto) problems with non-Boolean attributes. We propose the Choice Perceptron, a\nPerceptron-like algorithm for learning user preferences from set-wise choice\nfeedback over constructive domains and hybrid Boolean-numeric feature spaces.\nWe provide a theoretical analysis on the attained regret that holds for a large\nclass of query selection strategies, and devise a heuristic strategy that aims\nat optimizing the regret in practice. Finally, we demonstrate its effectiveness\nby empirical evaluation against existing competitors on constructive scenarios\nof increasing complexity.\n",
"title": "Constructive Preference Elicitation over Hybrid Combinatorial Spaces"
}
| null | null | null | null | true | null |
12948
| null |
Default
| null | null |
null |
{
"abstract": " We study the Ginzburg-Landau equations on Riemann surfaces of arbitrary\ngenus. In particular: we explicitly construct the (local moduli space of\ngauge-equivalent) solutions in a neighbourhood of a constant curvature branch\nof solutions; in linearizing the problem, we find a relation with de Rham\ncohomology groups of the surface; we classify holomorphic structures on line\nbundles arising as solutions to the equations in terms of the degree, the\nAbel-Jacobi map, and symmetric products of the surface; we construct explicitly\nthe automorphy factors and the equivariant connection on the trivial bundle\nover the Poincaré upper complex half plane.\n",
"title": "Ginzburg-Landau equations on Riemann surfaces of higher genus"
}
| null | null | null | null | true | null |
12949
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, a new class of frequency hopping sequences (FHSs) of length $\np^{n} $ is constructed by using Ding-Helleseth generalized cyclotomic classes\nof order 2, of which the Hamming auto- and cross-correlation functions are\ninvestigated (for the Hamming cross-correlation, only the case $ p\\equiv 3\\pmod\n4 $ is considered). It is shown that the set of the constructed FHSs is optimal\nwith respect to the average Hamming correlation functions.\n",
"title": "On the Hamming Auto- and Cross-correlation Functions of a Class of Frequency Hopping Sequences of Length $ p^{n} $"
}
| null | null | null | null | true | null |
12950
| null |
Default
| null | null |
null |
{
"abstract": " Scikit-multiflow is a multi-output/multi-label and stream data mining\nframework for the Python programming language. Conceived to serve as a platform\nto encourage democratization of stream learning research, it provides multiple\nstate of the art methods for stream learning, stream generators and evaluators.\nscikit-multiflow builds upon popular open source frameworks including\nscikit-learn, MOA and MEKA. Development follows the FOSS principles and quality\nis enforced by complying with PEP8 guidelines and using continuous integration\nand automatic testing. The source code is publicly available at\nthis https URL.\n",
"title": "Scikit-Multiflow: A Multi-output Streaming Framework"
}
| null | null | null | null | true | null |
12951
| null |
Default
| null | null |
null |
{
"abstract": " Distributed computing platforms provide a robust mechanism to perform\nlarge-scale computations by splitting the task and data among multiple\nlocations, possibly located thousands of miles apart geographically. Although\nsuch distribution of resources can lead to benefits, it also comes with its\nassociated problems such as rampant duplication of file transfers increasing\ncongestion, long job completion times, unexpected site crashing, suboptimal\ndata transfer rates, unpredictable reliability in a time range, and suboptimal\nusage of storage elements. In addition, each sub-system becomes a potential\nfailure node that can trigger system wide disruptions. In this vision paper, we\noutline our approach to leveraging Deep Learning algorithms to discover\nsolutions to unique problems that arise in a system with computational\ninfrastructure that is spread over a wide area. The presented vision, motivated\nby a real scientific use case from Belle II experiments, is to develop\nmultilayer neural networks to tackle forecasting, anomaly detection and\noptimization challenges in a complex and distributed data movement environment.\nThrough this vision based on Deep Learning principles, we aim to achieve\nreduced congestion events, faster file transfer rates, and enhanced site\nreliability.\n",
"title": "Deep Learning on Operational Facility Data Related to Large-Scale Distributed Area Scientific Workflows"
}
| null | null |
[
"Computer Science"
] | null | true | null |
12952
| null |
Validated
| null | null |
null |
{
"abstract": " We prove that the sectional category of the universal fibration with fibre X,\nfor X any space that satisfies a well-known conjecture of Halperin, equals one\nafter rationalization.\n",
"title": "The Rational Sectional Category of Certain Universal Fibrations"
}
| null | null |
[
"Mathematics"
] | null | true | null |
12953
| null |
Validated
| null | null |
null |
{
"abstract": " Planetary rings produce a distinct shape distortion in transit lightcurves.\nHowever, to accurately model such lightcurves the observations need to cover\nthe entire transit, especially ingress and egress, as well as an out-of-transit\nbaseline. Such observations can be challenging for long period planets, where\nthe transits may last for over a day. Planetary rings will also impact the\nshape of absorption lines in the stellar spectrum, as the planet and rings\ncover different parts of the rotating star (the Rossiter-McLaughlin effect).\nThese line-profile distortions depend on the size, structure, opacity,\nobliquity and sky projected angle of the ring system. For slow rotating stars,\nthis mainly impacts the amplitude of the induced velocity shift, however, for\nfast rotating stars the large velocity gradient across the star allows the line\ndistortion to be resolved, enabling direct determination of the ring\nparameters. We demonstrate that by modeling these distortions we can recover\nring system parameters (sky-projected angle, obliquity and size) using only a\nsmall part of the transit. Substructure in the rings, e.g. gaps, can be\nrecovered if the width of the features ($\\delta W$) relative to the size of the\nstar is similar to the intrinsic velocity resolution (set by the width of the\nlocal stellar profile, $\\gamma$) relative to the stellar rotation velocity ($v$\nsin$i$, i.e. $\\delta W / R_* \\gtrsim v$sin$i$/$\\gamma$). This opens up a new\nway to study the ring systems around planets with long orbital periods, where\nobservations of the full transit, covering the ingress and egress, are not\nalways feasible.\n",
"title": "Characterising exo-ringsystems around fast-rotating stars using the Rossiter-McLaughlin effect"
}
| null | null | null | null | true | null |
12954
| null |
Default
| null | null |
null |
{
"abstract": " We introduce a Unified Disentanglement Network (UFDN) trained on The Cancer\nGenome Atlas (TCGA). We demonstrate that the UFDN learns a biologically\nrelevant latent space of gene expression data by applying our network to two\nclassification tasks of cancer status and cancer type. Our UFDN specific\nalgorithms perform comparably to random forest methods. The UFDN allows for\ncontinuous, partial interpolation between distinct cancer types. Furthermore,\nwe perform an analysis of differentially expressed genes between skin cutaneous\nmelanoma(SKCM) samples and the same samples interpolated into glioblastoma\n(GBM). We demonstrate that our interpolations learn relevant metagenes that\nrecapitulate known glioblastoma mechanisms and suggest possible starting points\nfor investigations into the metastasis of SKCM into GBM.\n",
"title": "Learning a Generative Model of Cancer Metastasis"
}
| null | null | null | null | true | null |
12955
| null |
Default
| null | null |
null |
{
"abstract": " This paper provides a unified framework to deal with the challenges arising\nin dense cloud radio access networks (C-RAN), which include huge power\nconsumption, limited fronthaul capacity, heavy computational complexity,\nunavailability of full channel state information (CSI), etc. Specifically, we\naim to jointly optimize the remote radio head (RRH) selection, user equipment\n(UE)-RRH associations and beam-vectors to minimize the total network power\nconsumption (NPC) for dense multi-channel downlink C-RAN with incomplete CSI\nsubject to per-RRH power constraints, each UE's total rate requirement, and\nfronthaul link capacity constraints. This optimization problem is NP-hard. In\naddition, due to the incomplete CSI, the exact expression of UEs' rate\nexpression is intractable. We first conservatively replace UEs' rate expression\nwith its lower-bound. Then, based on the successive convex approximation (SCA)\ntechnique and the relationship between the data rate and the mean square error\n(MSE), we propose a single-layer iterative algorithm to solve the NPC\nminimization problem with convergence guarantee. In each iteration of the\nalgorithm, the Lagrange dual decomposition method is used to derive the\nstructure of the optimal beam-vectors, which facilitates the parallel\ncomputations at the Baseband unit (BBU) pool. Furthermore, a bisection UE\nselection algorithm is proposed to guarantee the feasibility of the problem.\nSimulation results show the benefits of the proposed algorithms and the fact\nthat a limited amount of CSI is sufficient to achieve performance close to that\nobtained when perfect CSI is possessed.\n",
"title": "Joint User Selection and Energy Minimization for Ultra-Dense Multi-channel C-RAN with Incomplete CSI"
}
| null | null |
[
"Computer Science"
] | null | true | null |
12956
| null |
Validated
| null | null |
null |
{
"abstract": " We study a class of determinant inequalities that are closely related to\nSidorenko's famous conjecture (Also conjectured by Erd\\H os and Simonovits in a\ndifferent form). Our results can also be interpreted as entropy inequalities\nfor Gaussian Markov random fields (GMRF). We call a GMRF on a finite graph $G$\nhomogeneous if the marginal distributions on the edges are all identical. We\nshow that if $G$ satisfies Sidorenko's conjecture then the differential entropy\nof any homogeneous GMRF on $G$ is at least $|E(G)|$ times the edge entropy plus\n$|V(G)|-2|E(G)|$ times the point entropy. We also prove this inequality in a\nlarge class of graphs for which Sidorenko's conjecture is not verified\nincluding the so-called Möbius ladder: $K_{5,5}\\setminus C_{10}$. The\nconnection between Sidorenko's conjecture and GMRF's is established via a large\ndeviation principle on high dimensional spheres combined with graph limit\ntheory.\n",
"title": "On Sidorenko's conjecture for determinants and Gaussian Markov random fields"
}
| null | null |
[
"Mathematics"
] | null | true | null |
12957
| null |
Validated
| null | null |
null |
{
"abstract": " Power grids are critical infrastructure assets that face non-technical losses\n(NTL) such as electricity theft or faulty meters. NTL may range up to 40% of\nthe total electricity distributed in emerging countries. Industrial NTL\ndetection systems are still largely based on expert knowledge when deciding\nwhether to carry out costly on-site inspections of customers. Electricity\nproviders are reluctant to move to large-scale deployments of automated systems\nthat learn NTL profiles from data due to the latter's propensity to suggest a\nlarge number of unnecessary inspections. In this paper, we propose a novel\nsystem that combines automated statistical decision making with expert\nknowledge. First, we propose a machine learning framework that classifies\ncustomers into NTL or non-NTL using a variety of features derived from the\ncustomers' consumption data. The methodology used is specifically tailored to\nthe level of noise in the data. Second, in order to allow human experts to feed\ntheir knowledge in the decision loop, we propose a method for visualizing\nprediction results at various granularity levels in a spatial hologram. Our\napproach allows domain experts to put the classification results into the\ncontext of the data and to incorporate their knowledge for making the final\ndecisions of which customers to inspect. This work has resulted in appreciable\nresults on a real-world data set of 3.6M customers. Our system is being\ndeployed in a commercial NTL detection software.\n",
"title": "Identifying Irregular Power Usage by Turning Predictions into Holographic Spatial Visualizations"
}
| null | null | null | null | true | null |
12958
| null |
Default
| null | null |
null |
{
"abstract": " Many recent studies of the motor system are divided into two distinct\napproaches: Those that investigate how motor responses are encoded in cortical\nneurons' firing rate dynamics and those that study the learning rules by which\nmammals and songbirds develop reliable motor responses. Computationally, the\nfirst approach is encapsulated by reservoir computing models, which can learn\nintricate motor tasks and produce internal dynamics strikingly similar to those\nof motor cortical neurons, but rely on biologically unrealistic learning rules.\nThe more realistic learning rules developed by the second approach are often\nderived for simplified, discrete tasks in contrast to the intricate dynamics\nthat characterize real motor responses. We bridge these two approaches to\ndevelop a biologically realistic learning rule for reservoir computing. Our\nalgorithm learns simulated motor tasks on which previous reservoir computing\nalgorithms fail, and reproduces experimental findings including those that\nrelate motor learning to Parkinson's disease and its treatment.\n",
"title": "A model of reward-modulated motor learning with parallelcortical and basal ganglia pathways"
}
| null | null |
[
"Quantitative Biology"
] | null | true | null |
12959
| null |
Validated
| null | null |
null |
{
"abstract": " The effective field theory of dark energy and modified gravity is supposed to\nwell describe, at low energies, the behaviour of the gravity modifications due\nto one extra scalar degree of freedom. The usual curvature perturbation is very\nuseful when studying the conditions for the avoidance of ghost instabilities as\nwell as the positivity of the squared speeds of propagation for both the scalar\nand tensor modes, or the Stückelberg field performs perfectly when\ninvestigating the evolution of linear perturbations. We show that the viable\nparameters space identified by requiring no-ghost instabilities and positive\nsquared speeds of propagation does not change by performing a field\nredefinition, while the requirement of the avoidance of tachyonic instability\nmight instead be different. Therefore, we find interesting to associate to the\ngeneral modified gravity theory described in the effective field theory\nframework, a perturbation field which will inherit the whole properties of the\ntheory. In the present paper we address the following questions: 1) how can we\ndefine such a field? and 2) what is the mass of such a field as the background\napproaches a final de Sitter state? We define a gauge invariant quantity which\nidentifies the density of the dark energy perturbation field valid for any\nbackground. We derive the mass associated to the gauge invariant dark energy\nfield on a de Sitter background, which we retain to be still a good\napproximation also at very low redshift ($z\\simeq 0$). On this background we\nalso investigate the value of the speed of propagation and we find that there\nexist classes of theories which admit a non-vanishing speed of propagation,\neven among the Horndeski model, for which in literature it has previously been\nfound a zero speed. We finally apply our results to specific well known models.\n",
"title": "A de Sitter limit analysis for dark energy and modified gravity models"
}
| null | null | null | null | true | null |
12960
| null |
Default
| null | null |
null |
{
"abstract": " Process-Aware Information Systems (PAIS) is an IT system that support\nbusiness processes and generate large amounts of event logs from the execution\nof business processes. An event log is represented as a tuple of CaseID,\nTimestamp, Activity and Actor. Process Mining is a new and emerging field that\naims at analyzing the event logs to discover, enhance and improve business\nprocesses and check conformance between run time and design time business\nprocesses. The large volume of event logs generated are stored in the\ndatabases. Relational databases perform well for a certain class of\napplications. However, there are a certain class of applications for which\nrelational databases are not able to scale. To handle such class of\napplications, NoSQL database systems emerged. Discovering a process model\n(workflow model) from event logs is one of the most challenging and important\nProcess Mining task. The $\\alpha$-miner algorithm is one of the first and most\nwidely used Process Discovery technique. Our objective is to investigate which\nof the databases (Relational or NoSQL) performs better for a Process Discovery\napplication under Process Mining. We implement the $\\alpha$-miner algorithm on\nrelational (row-oriented) and NoSQL (column-oriented) databases in database\nquery languages so that our algorithm is tightly coupled to the database. We\npresent a performance benchmarking and comparison of the $\\alpha$-miner\nalgorithm on row-oriented database and NoSQL column-oriented database so that\nwe can compare which database can efficiently store massive event logs and\nanalyze it in seconds to discover a process model.\n",
"title": "Empirical Analysis on Comparing the Performance of Alpha Miner Algorithm in SQL Query Language and NoSQL Column-Oriented Databases Using Apache Phoenix"
}
| null | null | null | null | true | null |
12961
| null |
Default
| null | null |
null |
{
"abstract": " DNN-based cross-modal retrieval has become a research hotspot, by which users\ncan search results across various modalities like image and text. However,\nexisting methods mainly focus on the pairwise correlation and reconstruction\nerror of labeled data. They ignore the semantically similar and dissimilar\nconstraints between different modalities, and cannot take advantage of\nunlabeled data. This paper proposes Cross-modal Deep Metric Learning with\nMulti-task Regularization (CDMLMR), which integrates quadruplet ranking loss\nand semi-supervised contrastive loss for modeling cross-modal semantic\nsimilarity in a unified multi-task learning architecture. The quadruplet\nranking loss can model the semantically similar and dissimilar constraints to\npreserve cross-modal relative similarity ranking information. The\nsemi-supervised contrastive loss is able to maximize the semantic similarity on\nboth labeled and unlabeled data. Compared to the existing methods, CDMLMR\nexploits not only the similarity ranking information but also unlabeled\ncross-modal data, and thus boosts cross-modal retrieval accuracy.\n",
"title": "Cross-modal Deep Metric Learning with Multi-task Regularization"
}
| null | null | null | null | true | null |
12962
| null |
Default
| null | null |
null |
{
"abstract": " In this survey paper, we give an overview of our recent works on the study of\nthe $W$-entropy for the heat equation associated with the Witten Laplacian on\nsuper-Ricci flows and the Langevin deformation on Wasserstein space over\nRiemannian manifolds. Inspired by Perelman's seminal work on the entropy\nformula for the Ricci flow, we prove the $W$-entropy formula for the heat\nequation associated with the Witten Laplacian on $n$-dimensional complete\nRiemannian manifolds with the $CD(K, m)$-condition, and the $W$-entropy formula\nfor the heat equation associated with the time dependent Witten Laplacian on\n$n$-dimensional compact manifolds equipped with a $(K, m)$-super Ricci flow,\nwhere $K\\in \\mathbb{R}$ and $m\\in [n, \\infty]$. Furthermore, we prove an\nanalogue of the $W$-entropy formula for the geodesic flow on the Wasserstein\nspace over Riemannian manifolds. Our result recaptures an important result due\nto Lott and Villani on the displacement convexity of the Boltzmann-Shannon\nentropy on Riemannian manifolds with non-negative Ricci curvature. To better\nunderstand the similarity between above two $W$-entropy formulas, we introduce\nthe Langevin deformation of geometric flows on the cotangent bundle over the\nWasserstein space and prove an extension of the $W$-entropy formula for the\nLangevin deformation. Finally, we make a discussion on the $W$-entropy for the\nRicci flow from the point of view of statistical mechanics and probability\ntheory.\n",
"title": "$W$-entropy formulas on super Ricci flows and Langevin deformation on Wasserstein space over Riemannian manifolds"
}
| null | null | null | null | true | null |
12963
| null |
Default
| null | null |
null |
{
"abstract": " We completely determine all commutative semigroup varieties that are\ncancellable elements of the lattice SEM of all semigroup varieties. In\nparticular, we prove that, for commutative varieties, the properties of being\ncancellable and modular elements of SEM are equivalent.\n",
"title": "Cancellable elements of the lattice of semigroup varieties"
}
| null | null | null | null | true | null |
12964
| null |
Default
| null | null |
null |
{
"abstract": " The availability of large scale event data with time stamps has given rise to\ndynamically evolving knowledge graphs that contain temporal information for\neach edge. Reasoning over time in such dynamic knowledge graphs is not yet well\nunderstood. To this end, we present Know-Evolve, a novel deep evolutionary\nknowledge network that learns non-linearly evolving entity representations over\ntime. The occurrence of a fact (edge) is modeled as a multivariate point\nprocess whose intensity function is modulated by the score for that fact\ncomputed based on the learned entity embeddings. We demonstrate significantly\nimproved performance over various relational learning approaches on two large\nscale real-world datasets. Further, our method effectively predicts occurrence\nor recurrence time of a fact which is novel compared to prior reasoning\napproaches in multi-relational setting.\n",
"title": "Know-Evolve: Deep Temporal Reasoning for Dynamic Knowledge Graphs"
}
| null | null |
[
"Computer Science"
] | null | true | null |
12965
| null |
Validated
| null | null |
null |
{
"abstract": " Deep neural networks are widely used in various domains. However, the nature\nof computations at each layer of the deep networks is far from being well\nunderstood. Increasing the interpretability of deep neural networks is thus\nimportant. Here, we construct a mean-field framework to understand how compact\nrepresentations are developed across layers, not only in deterministic deep\nnetworks with random weights but also in generative deep networks where an\nunsupervised learning is carried out. Our theory shows that the deep\ncomputation implements a dimensionality reduction while maintaining a finite\nlevel of weak correlations between neurons for possible feature extraction.\nMechanisms of dimensionality reduction and decorrelation are unified in the\nsame framework. This work may pave the way for understanding how a sensory\nhierarchy works.\n",
"title": "Mechanisms of dimensionality reduction and decorrelation in deep neural networks"
}
| null | null | null | null | true | null |
12966
| null |
Default
| null | null |
null |
{
"abstract": " Detecting strong ties among users in social and information networks is a\nfundamental operation that can improve performance on a multitude of\npersonalization and ranking tasks. Strong-tie edges are often readily obtained\nfrom the social network as users often participate in multiple overlapping\nnetworks via features such as following and messaging. These networks may vary\ngreatly in size, density and the information they carry. This setting leads to\na natural strong tie detection task: given a small set of labeled strong tie\nedges, how well can one detect unlabeled strong ties in the remainder of the\nnetwork?\nThis task becomes particularly daunting for the Twitter network due to scant\navailability of pairwise relationship attribute data, and sparsity of strong\ntie networks such as phone contacts. Given these challenges, a natural approach\nis to instead use structural network features for the task, produced by {\\em\ncombining} the strong and \"weak\" edges. In this work, we demonstrate via\nexperiments on Twitter data that using only such structural network features is\nsufficient for detecting strong ties with high precision. These structural\nnetwork features are obtained from the presence and frequency of small network\nmotifs on combined strong and weak ties. We observe that using motifs larger\nthan triads alleviate sparsity problems that arise for smaller motifs, both due\nto increased combinatorial possibilities as well as benefiting strongly from\nsearching beyond the ego network. Empirically, we observe that not all motifs\nare equally useful, and need to be carefully constructed from the combined\nedges in order to be effective for strong tie detection. Finally, we reinforce\nour experimental findings with providing theoretical justification that\nsuggests why incorporating these larger sized motifs as features could lead to\nincreased performance in planted graph models.\n",
"title": "Detecting Strong Ties Using Network Motifs"
}
| null | null | null | null | true | null |
12967
| null |
Default
| null | null |
null |
{
"abstract": " In many machine learning applications, it is important to explain the\npredictions of a black-box classifier. For example, why does a deep neural\nnetwork assign an image to a particular class? We cast interpretability of\nblack-box classifiers as a combinatorial maximization problem and propose an\nefficient streaming algorithm to solve it subject to cardinality constraints.\nBy extending ideas from Badanidiyuru et al. [2014], we provide a constant\nfactor approximation guarantee for our algorithm in the case of random stream\norder and a weakly submodular objective function. This is the first such\ntheoretical guarantee for this general class of functions, and we also show\nthat no such algorithm exists for a worst case stream order. Our algorithm\nobtains similar explanations of Inception V3 predictions $10$ times faster than\nthe state-of-the-art LIME framework of Ribeiro et al. [2016].\n",
"title": "Streaming Weak Submodularity: Interpreting Neural Networks on the Fly"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
12968
| null |
Validated
| null | null |
null |
{
"abstract": " We present high energy X-ray diffraction studies on the structural phases of\nan optimal high-$T_c$ superconductor La$_{2-x}$Sr$_x$CuO$_{4+y}$ tailored by\nco-hole-doping. This is specifically done by varying the content of two very\ndifferent chemical species, Sr and O, respectively, in order to study the\ninfluence of each. A superstructure known as staging is observed in all\nsamples, with the staging number $n$ increasing for higher Sr dopings $x$. We\nfind that the staging phases emerge abruptly with temperature, and can be\ndescribed as a second order phase transition with transition temperatures\nslightly depending on the Sr doping. The Sr appears to correlate the\ninterstitial oxygen in a way that stabilises the reproducibility of the staging\nphase both in terms of staging period and volume fraction in a specific sample.\nThe structural details as investigated in this letter appear to have no direct\nbearing on the electronic phase separation previously observed in the same\nsamples. This provides new evidence that the electronic phase separation is\ndetermined by the overall hole concentration rather than specific Sr/O content\nand concommittant structural details.\n",
"title": "Staging superstructures in high-$T_c$ Sr/O co-doped La$_{2-x}$Sr$_x$CuO$_{4+y}$"
}
| null | null | null | null | true | null |
12969
| null |
Default
| null | null |
null |
{
"abstract": " We have studied the structural, electronic and magnetic properties of spinel\n$\\rm Co_3O_4$(111) surfaces and their interfaces with ZnO (0001) using density\nfunctional theory (DFT) within the Generalized Gradient Approximation with\non-site Coulomb repulsion term (GGA+U). Two possible forms of spinel surface,\ncontaining $\\rm Co^{2+} $ and $\\rm Co^{3+} $ ions and terminated with either\ncobalt or oxygen ions were considered, as well as their interface with zinc\noxide. Our calculations demonstrate that $\\rm Co^{3+} $ ions attain non-zero\nmagnetic moments at the surface and interface, in contrast to the bulk, where\nthey are not magnetic, leading to the ferromagnetic ordering. Since heavily\nCo-doped ZnO samples can contain $\\rm Co_3O_4 $ secondary phase, such a\nmagnetic ordering at the interface might explain the origin of the magnetism in\nthese diluted magnetic semiconductors (DMS).\n",
"title": "Interface magnetism and electronic structure: ZnO(0001)/Co3O4(111)"
}
| null | null | null | null | true | null |
12970
| null |
Default
| null | null |
null |
{
"abstract": " Let $\\mathbb{B}$ be the unit ball of a complex Banach space $X$. In this\npaper, we will generalize the Bloch-type spaces and the little Bloch-type\nspaces to the open unit ball $\\mathbb{B}$ by using the radial derivative. Next,\nwe define an extended Cesàro operator $T_{\\varphi}$ with holomorphic symbol\n$\\varphi$ and characterize those $\\varphi$ for which $T_{\\varphi}$ is bounded\nbetween the Bloch-type spaces and the little Bloch-type spaces. We also\ncharacterize those $\\varphi$ for which $T_{\\varphi}$ is compact between the\nBloch-type spaces and the little Bloch-type spaces under some additional\nassumption on the symbol $\\varphi$. When $\\mathbb{B}$ is the open unit ball of\na finite dimensional complex Banach space $X$, this additional assumption is\nautomatically satisfied.\n",
"title": "Bloch-type spaces and extended Cesàro operators in the unit ball of a complex Banach space"
}
| null | null | null | null | true | null |
12971
| null |
Default
| null | null |
null |
{
"abstract": " The Apache Spark framework for distributed computation is popular in the data\nanalytics community due to its ease of use, but its MapReduce-style programming\nmodel can incur significant overheads when performing computations that do not\nmap directly onto this model. One way to mitigate these costs is to off-load\ncomputations onto MPI codes. In recent work, we introduced Alchemist, a system\nfor the analysis of large-scale data sets. Alchemist calls MPI-based libraries\nfrom within Spark applications, and it has minimal coding, communication, and\nmemory overheads. In particular, Alchemist allows users to retain the\nproductivity benefits of working within the Spark software ecosystem without\nsacrificing performance efficiency in linear algebra, machine learning, and\nother related computations.\nIn this paper, we discuss the motivation behind the development of Alchemist,\nand we provide a detailed overview its design and usage. We also demonstrate\nthe efficiency of our approach on medium-to-large data sets, using some\nstandard linear algebra operations, namely matrix multiplication and the\ntruncated singular value decomposition of a dense matrix, and we compare the\nperformance of Spark with that of Spark+Alchemist. These computations are run\non the NERSC supercomputer Cori Phase 1, a Cray XC40.\n",
"title": "Alchemist: An Apache Spark <=> MPI Interface"
}
| null | null | null | null | true | null |
12972
| null |
Default
| null | null |
null |
{
"abstract": " This paper is concerned with a compositional approach for constructing both\ninfinite (reduced-order models) and finite abstractions (a.k.a. finite Markov\ndecision processes) of large-scale interconnected discrete-time stochastic\ncontrol systems. The proposed framework is based on the notion of stochastic\nsimulation functions enabling us to use an abstract system as a substitution of\nthe original one in the controller design process with guaranteed error bounds.\nIn the first part of the paper, we derive sufficient small-gain type conditions\nfor the compositional quantification of the probabilistic distance between the\ninterconnection of stochastic control subsystems and that of their infinite\nabstractions. We then construct infinite abstractions together with their\ncorresponding stochastic simulation functions for a class of discrete-time\nnonlinear stochastic control systems. In the second part of the paper, we\nleverage small-gain type conditions for the compositional construction of\nfinite abstractions. We propose an approach to construct finite Markov decision\nprocesses (MDPs) of the concrete models (or their reduced-order versions)\nsatisfying an incremental input-to-state stability property. We also show that\nfor a particular class of nonlinear stochastic control systems, the\naforementioned property can be readily checked by matrix inequalities. We\ndemonstrate the effectiveness of the proposed results by applying our\napproaches to the temperature regulation in a circular building and\nconstructing compositionally a finite abstraction of a network containing 1000\nrooms. We also apply our proposed techniques to a fully connected network of 20\nnonlinear subsystems (totally 100 dimensions) and construct finite MDPs from\ntheir reduced-order versions (together 20 dimensions) with guaranteed error\nbounds on their output trajectories.\n",
"title": "Compositional (In)Finite Abstractions for Large-Scale Interconnected Stochastic Systems"
}
| null | null |
[
"Computer Science"
] | null | true | null |
12973
| null |
Validated
| null | null |
null |
{
"abstract": " For the polynomial ring over an arbitrary field with twelve variables, there\nexists a prime ideal whose symbolic Rees algebra is not finitely generated.\n",
"title": "Infinitely generated symbolic Rees algebras over finite fields"
}
| null | null | null | null | true | null |
12974
| null |
Default
| null | null |
null |
{
"abstract": " Deep generative neural networks have proven effective at both conditional and\nunconditional modeling of complex data distributions. Conditional generation\nenables interactive control, but creating new controls often requires expensive\nretraining. In this paper, we develop a method to condition generation without\nretraining the model. By post-hoc learning latent constraints, value functions\nthat identify regions in latent space that generate outputs with desired\nattributes, we can conditionally sample from these regions with gradient-based\noptimization or amortized actor functions. Combining attribute constraints with\na universal \"realism\" constraint, which enforces similarity to the data\ndistribution, we generate realistic conditional images from an unconditional\nvariational autoencoder. Further, using gradient-based optimization, we\ndemonstrate identity-preserving transformations that make the minimal\nadjustment in latent space to modify the attributes of an image. Finally, with\ndiscrete sequences of musical notes, we demonstrate zero-shot conditional\ngeneration, learning latent constraints in the absence of labeled data or a\ndifferentiable reward function. Code with dedicated cloud instance has been\nmade publicly available (this https URL).\n",
"title": "Latent Constraints: Learning to Generate Conditionally from Unconditional Generative Models"
}
| null | null | null | null | true | null |
12975
| null |
Default
| null | null |
null |
{
"abstract": " Zinc oxide and Aluminum Ferrite were prepared Chemical route. The samples\nwere characterized by XRD and VSM. Simulation of M-H plots of Co/CoO thin films\nwere performed. Effect of parameters was observed on saturation magnetization.\n",
"title": "Characterization of Zinc oxide & Aluminum Ferrite and Simulation studies of M-H plots of Cobalt/Cobaltoxide"
}
| null | null | null | null | true | null |
12976
| null |
Default
| null | null |
null |
{
"abstract": " We introduce a notion of weakly log-canonical Poisson structures on positive\nvarieties with potentials. Such a Poisson structure is log-canonical up to\nterms dominated by the potential. To a compatible real form of a weakly\nlog-canonical Poisson variety we assign an integrable system on the product of\na certain real convex polyhedral cone (the tropicalization of the variety) and\na compact torus.\nWe apply this theory to the dual Poisson-Lie group $G^*$ of a\nsimply-connected semisimple complex Lie group $G$. We define a positive\nstructure and potential on $G^*$ and show that the natural Poisson-Lie\nstructure on $G^*$ is weakly log-canonical with respect to this positive\nstructure and potential.\nFor $K \\subset G$ the compact real form, we show that the real form $K^*\n\\subset G^*$ is compatible and prove that the corresponding integrable system\nis defined on the product of the decorated string cone and the compact torus of\ndimension $\\frac{1}{2}({\\rm dim} \\, G - {\\rm rank} \\, G)$.\n",
"title": "Poisson Structures and Potentials"
}
| null | null | null | null | true | null |
12977
| null |
Default
| null | null |
null |
{
"abstract": " If $X$ is a compact Hausdorff space and $\\sigma$ is a homeomorphism of $X$,\nthen an involutive Banach algebra $\\ell^1(\\Sigma)$ of crossed product type is\nnaturally associated with the topological dynamical system $\\Sigma=(X,\\sigma)$.\nWe initiate the study of the relation between two-sided ideals of\n$\\ell^1(\\Sigma)$ and ${\\mathrm C}^\\ast(\\Sigma)$, the enveloping\n$\\mathrm{C}^\\ast$-algebra ${\\mathrm C}(X)\\rtimes_\\sigma \\mathbb Z$ of\n$\\ell^1(\\Sigma)$. Among others, we prove that the closure of a proper two-sided\nideal of $\\ell^1(\\Sigma)$ in ${\\mathrm C}^\\ast(\\Sigma)$ is again a proper\ntwo-sided ideal of ${\\mathrm C}^\\ast(\\Sigma)$.\n",
"title": "The closure of ideals of $\\boldsymbol{\\ell^1(Σ)}$ in its enveloping $\\boldsymbol{\\mathrm{C}^\\ast}$-algebra"
}
| null | null | null | null | true | null |
12978
| null |
Default
| null | null |
null |
{
"abstract": " The classical quadratic formula and some of its lesser known variants for\nsolving the quadratic equation are reviewed. Then, a new formula for the roots\nof a quadratic polynomial is presented.\n",
"title": "An alternative quadratic formula"
}
| null | null | null | null | true | null |
12979
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we present a gated convolutional recurrent neural network\nbased approach to solve task 4, large-scale weakly labelled semi-supervised\nsound event detection in domestic environments, of the DCASE 2018 challenge.\nGated linear units and a temporal attention layer are used to predict the onset\nand offset of sound events in 10s long audio clips. Whereby for training only\nweakly-labelled data is used. Virtual adversarial training is used for\nregularization, utilizing both labelled and unlabeled data. Furthermore, we\nintroduce self-adaptive label refinement, a method which allows unsupervised\nadaption of our trained system to refine the accuracy of frame-level class\npredictions. The proposed system reaches an overall macro averaged event-based\nF-score of 34.6%, resulting in a relative improvement of 20.5% over the\nbaseline system.\n",
"title": "Sound event detection using weakly-labeled semi-supervised data with GCRNNS, VAT and Self-Adaptive Label Refinement"
}
| null | null | null | null | true | null |
12980
| null |
Default
| null | null |
null |
{
"abstract": " We study flat FLRW $\\alpha$-attractor $\\mathrm{E}$- and $\\mathrm{T}$-models\nby introducing a dynamical systems framework that yields regularized\nunconstrained field equations on two-dimensional compact state spaces. This\nresults in both illustrative figures and a complete description of the entire\nsolution spaces of these models, including asymptotics. In particular, it is\nshown that observational viability, which requires a sufficient number of\n$e$-folds, is associated with a solution given by a one-dimensional center\nmanifold of a past asymptotic de Sitter state, where the center manifold\nstructure also explains why nearby solutions are attracted to this\n`inflationary attractor solution.' A center manifold expansion yields a\ndescription of the inflationary regime with arbitrary analytic accuracy, where\nthe slow-roll approximation asymptotically describes the tangency condition of\nthe center manifold at the asymptotic de Sitter state.\n",
"title": "Inflationary $α$-attractor cosmology: A global dynamical systems perspective"
}
| null | null |
[
"Physics"
] | null | true | null |
12981
| null |
Validated
| null | null |
null |
{
"abstract": " Let $\\Omega\\subset\\mathbb{R}^{n+1}$ have minimal Gaussian surface area among\nall sets satisfying $\\Omega=-\\Omega$ with fixed Gaussian volume. Let $A=A_{x}$\nbe the second fundamental form of $\\partial\\Omega$ at $x$, i.e. $A$ is the\nmatrix of first order partial derivatives of the unit normal vector at\n$x\\in\\partial\\Omega$. For any $x=(x_{1},\\ldots,x_{n+1})\\in\\mathbb{R}^{n+1}$,\nlet $\\gamma_{n}(x)=(2\\pi)^{-n/2}e^{-(x_{1}^{2}+\\cdots+x_{n+1}^{2})/2}$. Let\n$\\|A\\|^{2}$ be the sum of the squares of the entries of $A$, and let\n$\\|A\\|_{2\\to 2}$ denote the $\\ell_{2}$ operator norm of $A$.\nIt is shown that if $\\Omega$ or $\\Omega^{c}$ is convex, and if either\n$$\\int_{\\partial\\Omega}(\\|A_{x}\\|^{2}-1)\\gamma_{n}(x)dx>0\\qquad\\mbox{or}\\qquad\n\\int_{\\partial\\Omega}\\Big(\\|A_{x}\\|^{2}-1+2\\sup_{y\\in\\partial\\Omega}\\|A_{y}\\|_{2\\to\n2}^{2}\\Big)\\gamma_{n}(x)dx<0,$$ then $\\partial\\Omega$ must be a round cylinder.\nThat is, except for the case that the average value of $\\|A\\|^{2}$ is slightly\nless than $1$, we resolve the convex case of a question of Barthe from 2001.\nThe main tool is the Colding-Minicozzi theory for Gaussian minimal surfaces,\nwhich studies eigenfunctions of the Ornstein-Uhlenbeck type operator $L=\n\\Delta-\\langle x,\\nabla \\rangle+\\|A\\|^{2}+1$ associated to the surface\n$\\partial\\Omega$. A key new ingredient is the use of a randomly chosen degree 2\npolynomial in the second variation formula for the Gaussian surface area. Our\nactual results are a bit more general than the above statement. Also, some of\nour results hold without the assumption of convexity.\n",
"title": "Symmetric Convex Sets with Minimal Gaussian Surface Area"
}
| null | null | null | null | true | null |
12982
| null |
Default
| null | null |
null |
{
"abstract": " Continuum Approximation (CA) is an efficient and parsimonious technique for\nmodeling complex logistics problems. In this paper,we review recent studies\nthat develop CA models for transportation, distribution and logistics problems\nwith the aim of synthesizing recent advancements and identifying current\nresearch gaps. This survey focuses on important principles and key results from\nCA models. In particular, we consider how these studies fill the gaps\nidentified by the most recent literature reviews in this field. We observe that\nCA models are used in a wider range of applications, especially in the areas of\nfacility location and integrated supply chain management. Most studies use CA\nas an alternative to exact solution approaches; however, CA can also be used in\ncombination with exact approaches. We also conclude with promising areas of\nfuture work.\n",
"title": "Advancements in Continuum Approximation Models for Logistics and Transportation Systems: 1996 - 2016"
}
| null | null | null | null | true | null |
12983
| null |
Default
| null | null |
null |
{
"abstract": " From minimal surfaces such as Simons' cone and catenoids, using refined\nLyapunov-Schmidt reduction method, we construct new solutions for a free\nboundary problem whose free boundary has two components. In dimension $8$,\nusing variational arguments, we also obtain solutions which are global\nminimizers of the corresponding energy functional. This shows that Savin's\ntheorem is optimal.\n",
"title": "On a free boundary problem and minimal surfaces"
}
| null | null | null | null | true | null |
12984
| null |
Default
| null | null |
null |
{
"abstract": " We first review traditional approaches to memory storage and formation,\ndrawing on the literature of quantitative neuroscience as well as statistical\nphysics. These have generally focused on the fast dynamics of neurons; however,\nthere is now an increasing emphasis on the slow dynamics of synapses, whose\nweight changes are held to be responsible for memory storage. An important\nfirst step in this direction was taken in the context of Fusi's cascade model,\nwhere complex synaptic architectures were invoked, in particular, to store\nlong-term memories. No explicit synaptic dynamics were, however, invoked in\nthat work. These were recently incorporated theoretically using the techniques\nused in agent-based modelling, and subsequently, models of competing and\ncooperating synapses were formulated. It was found that the key to the storage\nof long-term memories lay in the competitive dynamics of synapses. In this\nreview, we focus on models of synaptic competition and cooperation, and look at\nthe outstanding challenges that remain.\n",
"title": "Storing and retrieving long-term memories: cooperation and competition in synaptic dynamics"
}
| null | null |
[
"Quantitative Biology"
] | null | true | null |
12985
| null |
Validated
| null | null |
null |
{
"abstract": " Predictive modeling is increasingly being employed to assist human\ndecision-makers. One purported advantage of replacing or augmenting human\njudgment with computer models in high stakes settings-- such as sentencing,\nhiring, policing, college admissions, and parole decisions-- is the perceived\n\"neutrality\" of computers. It is argued that because computer models do not\nhold personal prejudice, the predictions they produce will be equally free from\nprejudice. There is growing recognition that employing algorithms does not\nremove the potential for bias, and can even amplify it if the training data\nwere generated by a process that is itself biased. In this paper, we provide a\nprobabilistic notion of algorithmic bias. We propose a method to eliminate bias\nfrom predictive models by removing all information regarding protected\nvariables from the data to which the models will ultimately be trained. Unlike\nprevious work in this area, our framework is general enough to accommodate data\non any measurement scale. Motivated by models currently in use in the criminal\njustice system that inform decisions on pre-trial release and parole, we apply\nour proposed method to a dataset on the criminal histories of individuals at\nthe time of sentencing to produce \"race-neutral\" predictions of re-arrest. In\nthe process, we demonstrate that a common approach to creating \"race-neutral\"\nmodels-- omitting race as a covariate-- still results in racially disparate\npredictions. We then demonstrate that the application of our proposed method to\nthese data removes racial disparities from predictions with minimal impact on\npredictive accuracy.\n",
"title": "An algorithm for removing sensitive information: application to race-independent recidivism prediction"
}
| null | null | null | null | true | null |
12986
| null |
Default
| null | null |
null |
{
"abstract": " Emotion cause extraction aims to identify the reasons behind a certain\nemotion expressed in text. It is a much more difficult task compared to emotion\nclassification. Inspired by recent advances in using deep memory networks for\nquestion answering (QA), we propose a new approach which considers emotion\ncause identification as a reading comprehension task in QA. Inspired by\nconvolutional neural networks, we propose a new mechanism to store relevant\ncontext in different memory slots to model context information. Our proposed\napproach can extract both word level sequence features and lexical features.\nPerformance evaluation shows that our method achieves the state-of-the-art\nperformance on a recently released emotion cause dataset, outperforming a\nnumber of competitive baselines by at least 3.01% in F-measure.\n",
"title": "A Question Answering Approach to Emotion Cause Extraction"
}
| null | null | null | null | true | null |
12987
| null |
Default
| null | null |
null |
{
"abstract": " Machine Learning on graph-structured data is an important and omnipresent\ntask for a vast variety of applications including anomaly detection and dynamic\nnetwork analysis. In this paper, a deep generative model is introduced to\ncapture continuous probability densities corresponding to the nodes of an\narbitrary graph. In contrast to all learning formulations in the area of\ndiscriminative pattern recognition, we propose a scalable generative\noptimization/algorithm theoretically proved to capture distributions at the\nnodes of a graph. Our model is able to generate samples from the probability\ndensities learned at each node. This probabilistic data generation model, i.e.\nconvolutional graph auto-encoder (CGAE), is devised based on the localized\nfirst-order approximation of spectral graph convolutions, deep learning, and\nthe variational Bayesian inference. We apply our CGAE to a new problem, the\nspatio-temporal probabilistic solar irradiance prediction. Multiple solar\nradiation measurement sites in a wide area in northern states of the US are\nmodeled as an undirected graph. Using our proposed model, the distribution of\nfuture irradiance given historical radiation observations is estimated for\nevery site/node. Numerical results on the National Solar Radiation Database\nshow state-of-the-art performance for probabilistic radiation prediction on\ngeographically distributed irradiance data in terms of reliability, sharpness,\nand continuous ranked probability score.\n",
"title": "Convolutional Graph Auto-encoder: A Deep Generative Neural Architecture for Probabilistic Spatio-temporal Solar Irradiance Forecasting"
}
| null | null | null | null | true | null |
12988
| null |
Default
| null | null |
null |
{
"abstract": " A metric space $X$ is quasisymmetrically co-Hopfian if every quasisymmetric\nembedding of $X$ into itself is onto. We construct the first examples of metric\nspaces homeomorphic to the universal Menger curve and higher dimensional\nSierpiński spaces, which are quasisymmetrically co-Hopfian. We also show that\nthe collection of quasisymmetric equivalence classes of spaces homeomorphic to\nthe Menger curve is uncountable. These results answer a problem and generalize\nresults of Merenkov from \\cite{Mer:coHopf}.\n",
"title": "Quasisymmetrically co-Hopfian Sierpiński Spaces and Menger Curve"
}
| null | null |
[
"Mathematics"
] | null | true | null |
12989
| null |
Validated
| null | null |
null |
{
"abstract": " We investigate a tight-binding electronic chain featuring diagonal and\noff-diagonal disorder, these being modelled through the long-range-correlated\nfractional Brownian motion. Particularly, by employing exact diagonalization\nmethods, we evaluate how the eigenstate spectrum of the system and its related\nsingle-particle dynamics respond to both competing sources of disorder.\nMoreover, we report the possibility of carrying out efficient end-to-end\nquantum-state transfer protocols even in the presence of such generalized\ndisorder due to the appearance of extended states around the middle of the band\nin the limit of strong correlations.\n",
"title": "Localization properties and high-fidelity state transfer in electronic hopping models with correlated disorder"
}
| null | null | null | null | true | null |
12990
| null |
Default
| null | null |
null |
{
"abstract": " We study revenue optimization learning algorithms for repeated posted-price\nauctions where a seller interacts with a single strategic buyer that holds a\nfixed private valuation for a good and seeks to maximize his cumulative\ndiscounted surplus. For this setting, first, we propose a novel algorithm that\nnever decreases offered prices and has a tight strategic regret bound in\n$\\Theta(\\log\\log T)$ under some mild assumptions on the buyer surplus\ndiscounting. This result closes the open research question on the existence of\na no-regret horizon-independent weakly consistent pricing. The proposed\nalgorithm is inspired by our observation that a double decrease of offered\nprices in a weakly consistent algorithm is enough to cause a linear regret.\nThis motivates us to construct a novel transformation that maps a\nright-consistent algorithm to a weakly consistent one that never decreases\noffered prices.\nSecond, we outperform the previously known strategic regret upper bound of\nthe algorithm PRRFES, where the improvement is achieved by means of a finer\nconstant factor $C$ of the principal term $C\\log\\log T$ in this upper bound.\nFinally, we generalize results on strategic regret previously known for\ngeometric discounting of the buyer's surplus to discounting of other types,\nnamely: the optimality of the pricing PRRFES to the case of geometrically\nconcave decreasing discounting; and linear lower bound on the strategic regret\nof a wide range of horizon-independent weakly consistent algorithms to the case\nof arbitrary discounts.\n",
"title": "On consistency of optimal pricing algorithms in repeated posted-price auctions with strategic buyer"
}
| null | null | null | null | true | null |
12991
| null |
Default
| null | null |
null |
{
"abstract": " The occurrence of drug-drug-interactions (DDI) from multiple drug\nprescriptions is a serious problem, both for individuals and health-care\nsystems, since patients with complications due to DDI are likely to re-enter\nthe system at a costlier level. We present a large-scale longitudinal study of\nthe DDI phenomenon at the primary- and secondary-care level using electronic\nhealth records from the city of Blumenau in Southern Brazil (pop. ~340,000).\nThis is the first study of DDI we are aware of that follows an entire city\nlongitudinally for 18 months. We found that 181 distinct drug pairs known to\ninteract were dispensed concomitantly to 12% of the patients in the city's\npublic health-care system. Further, 4% of the patients were dispensed major DDI\ncombinations, likely to result in very serious adverse reactions and costs we\nestimate to be larger than previously reported. DDI results are integrated into\nassociative networks for inference and visualization, revealing key medications\nand interactions. Analysis reveals that women have a 60% increased risk of DDI\nas compared to men; the increase becomes 90% when only major DDI are\nconsidered. Furthermore, DDI risk increases substantially with age. Patients\naged 70-79 years have a 34% risk of DDI when they are prescribed two or more\ndrugs concomitantly. Interestingly, a null model demonstrates that age and\nwomen-specific risks from increased polypharmacy far exceed expectations in\nthose populations. This suggests that social and biological factors are at\nplay. Finally, we demonstrate that machine learning classifiers accurately\npredict patients likely to be administered DDI given their history of\nprescribed drugs, gender, and age (MCC=.7,AUC=.97). These results demonstrate\nthat accurate warning systems for known DDI can be devised for health-care\nsystems leading to substantial reduction of DDI-related adverse reactions and\nhealth-care savings.\n",
"title": "City-wide Analysis of Electronic Health Records Reveals Gender and Age Biases in the Administration of Known Drug-Drug Interactions"
}
| null | null | null | null | true | null |
12992
| null |
Default
| null | null |
null |
{
"abstract": " In this article we study the stabilizing of a primitive pattern of behaviour\nfor the two-species community with chemotaxis due to the short-wavelength\nexternal signal. We use a system of Patlak-Keller-Segel type as a model of the\ncommunity. It is well-known that such systems can produce complex unsteady\npatterns of behaviour which are usually explained mathematically by\nbifurcations of some basic solutions that describe simpler patterns. As far as\nwe aware, all such bifurcations in the models of the Patlak-Keller-Segel type\nhad been found for homogeneous (i.e. translationally invariant) systems where\nthe basic solutions are equilibria with homogeneous distributions of all\nspecies. The model considered in the present paper does not possess the\ntranslational invariance: one of species (the predators) is assumed to be\ncapable of moving in response to a signal produced externally in addition to\nthe signal emitted by another species (the prey). For instance, the external\nsignal may arise from the inhomogeneity of the distribution of an environmental\ncharacteristic such as temperature, salinity, terrain relief, etc. Our goal is\nto examine the effect of short-wavelength inhomogeneity. To do this, we employ\na certain homogenization procedure. We separate the short-wavelength and smooth\ncomponents of the system response and derive a slow system governing the latter\none. Analysing the slow system and comparing it with the case of homogeneous\nenvironment shows that, generically, a short-wavelength inhomogeneity results\nin an exponential decrease in the motility of the predators. The loss of\nmotility prevents, to a great extent, the occurrence of complex unsteady\npatterns and dramatically stabilizes the primitive basic solution. In some\nsense, the necessity of dealing with intensive small-scale changes of the\nenvironment makes the system unable to respond to other challenges.\n",
"title": "A remark on the disorienting of species due to the fluctuating environment"
}
| null | null | null | null | true | null |
12993
| null |
Default
| null | null |
null |
{
"abstract": " We investigate some extremal problems in Fourier analysis and their\nconnection to a problem in prime number theory. In particular, we improve the\ncurrent bounds for the largest possible gap between consecutive primes assuming\nthe Riemann hypothesis.\n",
"title": "Fourier optimization and prime gaps"
}
| null | null |
[
"Mathematics"
] | null | true | null |
12994
| null |
Validated
| null | null |
null |
{
"abstract": " Knowledge bases are important resources for a variety of natural language\nprocessing tasks but suffer from incompleteness. We propose a novel embedding\nmodel, \\emph{ITransF}, to perform knowledge base completion. Equipped with a\nsparse attention mechanism, ITransF discovers hidden concepts of relations and\ntransfer statistical strength through the sharing of concepts. Moreover, the\nlearned associations between relations and concepts, which are represented by\nsparse attention vectors, can be interpreted easily. We evaluate ITransF on two\nbenchmark datasets---WN18 and FB15k for knowledge base completion and obtains\nimprovements on both the mean rank and Hits@10 metrics, over all baselines that\ndo not use additional information.\n",
"title": "An Interpretable Knowledge Transfer Model for Knowledge Base Completion"
}
| null | null | null | null | true | null |
12995
| null |
Default
| null | null |
null |
{
"abstract": " Gaussian processes (GPs) are highly flexible function estimators used for\ngeospatial analysis, nonparametric regression, and machine learning, but they\nare computationally infeasible for large datasets. Vecchia approximations of\nGPs have been used to enable fast evaluation of the likelihood for parameter\ninference. Here, we study Vecchia approximations of spatial predictions at\nobserved and unobserved locations, including obtaining joint predictive\ndistributions at large sets of locations. We propose a general Vecchia\nframework for GP predictions, which contains some novel and some existing\nspecial cases. We study the accuracy and computational properties of these\napproaches theoretically and numerically. We show that our new approaches\nexhibit linear computational complexity in the total number of spatial\nlocations. We also apply our methods to a satellite dataset of chlorophyll\nfluorescence.\n",
"title": "Vecchia approximations of Gaussian-process predictions"
}
| null | null | null | null | true | null |
12996
| null |
Default
| null | null |
null |
{
"abstract": " For the prediction with experts' advice setting, we consider some methods to\nconstruct forecasting algorithms that suffer loss not much more than any expert\nin the pool. In contrast to the standard approach, we investigate the case of\nlong-term forecasting of time series. This approach implies that each expert\nissues a forecast for a time point ahead (or a time interval), and then the\nmaster algorithm combines these forecasts into one aggregated forecast\n(sequence of forecasts). We introduce two new approaches to aggregating\nexperts' long-term interval predictions. Both are based on Vovk's aggregating\nalgorithm. The first approach applies the method of Mixing Past Posteriors\nmethod to the long-term prediction. The second approach is used for the\ninterval forecasting and considers overlapping experts. The upper bounds for\nregret of these algorithms for adversarial case are obtained. We also present\nthe results of numerical experiments on time series long-term prediction.\n",
"title": "Long-Term Sequential Prediction Using Expert Advice"
}
| null | null | null | null | true | null |
12997
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, a new adaptive multi-batch experience replay scheme is\nproposed for proximal policy optimization (PPO) for continuous action control.\nOn the contrary to original PPO, the proposed scheme uses the batch samples of\npast policies as well as the current policy for the update for the next policy,\nwhere the number of the used past batches is adaptively determined based on the\noldness of the past batches measured by the average importance sampling (IS)\nweight. The new algorithm constructed by combining PPO with the proposed\nmulti-batch experience replay scheme maintains the advantages of original PPO\nsuch as random mini-batch sampling and small bias due to low IS weights by\nstoring the pre-computed advantages and values and adaptively determining the\nmini-batch size. Numerical results show that the proposed method significantly\nincreases the speed and stability of convergence on various continuous control\ntasks compared to original PPO.\n",
"title": "AMBER: Adaptive Multi-Batch Experience Replay for Continuous Action Control"
}
| null | null | null | null | true | null |
12998
| null |
Default
| null | null |
null |
{
"abstract": " Muon-spin rotation data collected at ambient pressure ($p$) and at $p=2.42$\nGPa in MnP were analyzed to check their consistency with various low- and\nhigh-pressure magnetic structures reported in the literature. Our analysis\nconfirms that in MnP the low-temperature and low-pressure helimagnetic phase is\ncharacterised by an increased value of the average magnetic moment compared to\nthe high-temperature ferromagnetic phase. An elliptical double-helical\nstructure with a propagation vector ${\\bf Q}=(0,0,0.117)$, an $a-$axis moment\nelongated by approximately 18% and an additional tilt of the rotation plane\ntowards $c-$direction by $\\simeq 4-8^{\\rm o}$ leads to a good agreement between\nthe theory and the experiment. The analysis of the high-pressure $\\mu$SR data\nreveals that the new magnetic order appearing for pressures exceeding $1.5$ GPa\ncan not be described by keeping the propagation vector ${\\bf Q} \\parallel c$.\nEven the extreme case -- decoupling the double-helical structure into four\nindividual helices -- remains inconsistent with the experiment. It is shown\nthat the high-pressure magnetic phase which is a precursor of superconductivity\nis an incommensurate helical state with ${\\bf Q} \\parallel b$.\n",
"title": "Magnetic states of MnP: muon-spin rotation studies"
}
| null | null | null | null | true | null |
12999
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we describe the optical imaging data processing pipeline\ndeveloped for the Subaru Telescope's Hyper Suprime-Cam (HSC) instrument. The\nHSC Pipeline builds on the prototype pipeline being developed by the Large\nSynoptic Survey Telescope's Data Management system, adding customizations for\nHSC, large-scale processing capabilities, and novel algorithms that have since\nbeen reincorporated into the LSST codebase. While designed primarily to reduce\nHSC Subaru Strategic Program (SSP) data, it is also the recommended pipeline\nfor reducing general-observer HSC data. The HSC pipeline includes high level\nprocessing steps that generate coadded images and science-ready catalogs as\nwell as low-level detrending and image characterizations.\n",
"title": "The Hyper Suprime-Cam Software Pipeline"
}
| null | null | null | null | true | null |
13000
| null |
Default
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.