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": " Learning approaches have recently become very popular in the field of inverse\nproblems. A large variety of methods has been established in recent years,\nranging from bi-level learning to high-dimensional machine learning techniques.\nMost learning approaches, however, only aim at fitting parametrised models to\nfavourable training data whilst ignoring misfit training data completely. In\nthis paper, we follow up on the idea of learning parametrised regularisation\nfunctions by quotient minimisation as established in [3]. We extend the model\ntherein to include higher-dimensional filter functions to be learned and allow\nfor fit- and misfit-training data consisting of multiple functions. We first\npresent results resembling behaviour of well-established derivative-based\nsparse regularisers like total variation or higher-order total variation in\none-dimension. Our second and main contribution is the introduction of novel\nfamilies of non-derivative-based regularisers. This is accomplished by learning\nfavourable scales and geometric properties while at the same time avoiding\nunfavourable ones.\n",
"title": "Learning Filter Functions in Regularisers by Minimising Quotients"
}
| null | null |
[
"Mathematics"
] | null | true | null |
10901
| null |
Validated
| null | null |
null |
{
"abstract": " A general Boltzmann machine with continuous visible and discrete integer\nvalued hidden states is introduced. Under mild assumptions about the connection\nmatrices, the probability density function of the visible units can be solved\nfor analytically, yielding a novel parametric density function involving a\nratio of Riemann-Theta functions. The conditional expectation of a hidden state\nfor given visible states can also be calculated analytically, yielding a\nderivative of the logarithmic Riemann-Theta function. The conditional\nexpectation can be used as activation function in a feedforward neural network,\nthereby increasing the modelling capacity of the network. Both the Boltzmann\nmachine and the derived feedforward neural network can be successfully trained\nvia standard gradient- and non-gradient-based optimization techniques.\n",
"title": "Riemann-Theta Boltzmann Machine"
}
| null | null | null | null | true | null |
10902
| null |
Default
| null | null |
null |
{
"abstract": " We introduce a novel regression framework which simultaneously models the\nquantile and the Expected Shortfall (ES) of a response variable given a set of\ncovariates. This regression is based on a strictly consistent loss function for\nthe pair quantile and ES, which allows for M- and Z-estimation of the joint\nregression parameters. We show consistency and asymptotic normality for both\nestimators under weak regularity conditions. The underlying loss function\ndepends on two specification functions, whose choice affects the properties of\nthe resulting estimators. We find that the Z-estimator is numerically unstable\nand thus, we rely on M-estimation of the model parameters. Extensive\nsimulations verify the asymptotic properties and analyze the small sample\nbehavior of the M-estimator for different specification functions. This joint\nregression framework allows for various applications including estimating,\nforecasting, and backtesting ES, which is particularly relevant in light of the\nrecent introduction of ES into the Basel Accords.\n",
"title": "A Joint Quantile and Expected Shortfall Regression Framework"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
10903
| null |
Validated
| null | null |
null |
{
"abstract": " Representation learning has become an invaluable approach for learning from\nsymbolic data such as text and graphs. However, while complex symbolic datasets\noften exhibit a latent hierarchical structure, state-of-the-art methods\ntypically learn embeddings in Euclidean vector spaces, which do not account for\nthis property. For this purpose, we introduce a new approach for learning\nhierarchical representations of symbolic data by embedding them into hyperbolic\nspace -- or more precisely into an n-dimensional Poincaré ball. Due to the\nunderlying hyperbolic geometry, this allows us to learn parsimonious\nrepresentations of symbolic data by simultaneously capturing hierarchy and\nsimilarity. We introduce an efficient algorithm to learn the embeddings based\non Riemannian optimization and show experimentally that Poincaré embeddings\noutperform Euclidean embeddings significantly on data with latent hierarchies,\nboth in terms of representation capacity and in terms of generalization\nability.\n",
"title": "Poincaré Embeddings for Learning Hierarchical Representations"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
10904
| null |
Validated
| null | null |
null |
{
"abstract": " Polymer models are used to describe chromatin, which can be folded at\ndifferent spatial scales by binding molecules. By folding, chromatin generates\nloops of various sizes. We present here a randomly cross-linked (RCL) polymer\nmodel, where monomer pairs are connected randomly. We obtain asymptotic\nformulas for the steady-state variance, encounter probability, the radius of\ngyration, instantaneous displacement and the mean first encounter time between\nany two monomers. The analytical results are confirmed by Brownian simulations.\nFinally, the present results can be used to extract the minimum number of\ncross-links in a chromatin region from {conformation capture} data.\n",
"title": "Randomly cross-linked polymer models"
}
| null | null | null | null | true | null |
10905
| null |
Default
| null | null |
null |
{
"abstract": " Complex contagion models have been developed to understand a wide range of\nsocial phenomena such as adoption of cultural fads, the diffusion of belief,\nnorms, and innovations in social networks, and the rise of collective action to\njoin a riot. Most existing works focus on contagions where individuals' states\nare represented by {\\em binary} variables, and propagation takes place over a\nsingle isolated network. However, characterization of an individual's standing\non a given matter as a binary state might be overly simplistic as most of our\nopinions, feelings, and perceptions vary over more than two states. Also, most\nreal-world contagions take place over multiple networks (e.g., Twitter and\nFacebook) or involve {\\em multiplex} networks where individuals engage in\ndifferent {\\em types} of relationships (e.g., acquaintance, co-worker, family,\netc.). To this end, this paper studies {\\em multi-stage} complex contagions\nthat take place over multi-layer or multiplex networks. Under a linear\nthreshold based contagion model, we give analytic results for the probability\nand expected size of \\textit{global} cascades, i.e., cases where a randomly\nchosen node can initiate a propagation that eventually reaches a {\\em positive}\nfraction of the whole population. Analytic results are also confirmed and\nsupported by an extensive numerical study. In particular, we demonstrate how\nthe dynamics of complex contagions is affected by the extra weight exerted by\n\\textit{hyper-active} nodes and by the structural properties of the networks\ninvolved. Among other things, we reveal an interesting connection between the\nassortativity of a network and the impact of \\textit{hyper-active} nodes on the\ncascade size.\n",
"title": "Multi-Stage Complex Contagions in Random Multiplex Networks"
}
| null | null | null | null | true | null |
10906
| null |
Default
| null | null |
null |
{
"abstract": " In network coding, we discuss the effect of sequential error injection on\ninformation leakage. We show that there is no improvement when the operations\nin the network are linear operations. However, when the operations in the\nnetwork contains non-linear operations, we find a counterexample to improve\nEve's obtained information. Furthermore, we discuss the asymptotic rate in a\nlinear network under the secrecy and robustness conditions as well as under the\nsecrecy condition alone. Finally, we apply our results to network quantum key\ndistribution, which clarifies the type of network that enables us to realize\nsecure long distance communication via short distance quantum key distribution.\n",
"title": "Secrecy and Robustness for Active Attack in Secure Network Coding and its Application to Network Quantum Key Distribution"
}
| null | null | null | null | true | null |
10907
| null |
Default
| null | null |
null |
{
"abstract": " Simultaneous Localization and Mapping (SLAM) is the problem of constructing a\nmap of an agent's environment while localizing or tracking the mobile agent's\nposition and orientation within the map. Algorithms for SLAM have high\ncomputational requirements, which has hindered their use on embedded devices.\nApproximation can be used to reduce the time and energy requirements of SLAM\nimplementations as long as the approximations do not prevent the agent from\nnavigating correctly through the environment. Previous studies of approximation\nin SLAM have assumed that the entire trajectory of the agent is known before\nthe agent starts to move, and they have focused on offline controllers that use\nfeatures of the trajectory to set approximation knobs at the start of the\ntrajectory. In practice, the trajectory is not usually known ahead of time, and\nallowing knob settings to change dynamically opens up more opportunities for\nreducing computation time and energy.\nWe describe SLAMBooster, an application-aware online control system for SLAM\nthat adaptively controls approximation knobs during the motion of the agent.\nSLAMBooster is based on a control technique called hierarchical proportional\ncontrol but our experiments showed this application-agnostic control led to an\nunacceptable reduction in the quality of localization. To address this problem,\nSLAMBooster exploits domain knowledge: it uses features extracted from input\nframes and from the estimated motion of the agent in its algorithm for\ncontrolling approximation.\nWe implemented SLAMBooster in the open-source SLAMBench framework. Our\nexperiments show that SLAMBooster reduces the computation time and energy\nconsumption by around half on the average on an embedded platform, while\nmaintaining the accuracy of the localization within reasonable bounds. These\nimprovements make it feasible to deploy SLAM on a wider range of devices.\n",
"title": "SLAMBooster: An Application-aware Controller for Approximation in SLAM"
}
| null | null |
[
"Computer Science"
] | null | true | null |
10908
| null |
Validated
| null | null |
null |
{
"abstract": " Characterizing a patient's progression through stages of sepsis is critical\nfor enabling risk stratification and adaptive, personalized treatment. However,\ncommonly used sepsis diagnostic criteria fail to account for significant\nunderlying heterogeneity, both between patients as well as over time in a\nsingle patient. We introduce a hidden Markov model of sepsis progression that\nexplicitly accounts for patient heterogeneity. Benchmarked against two sepsis\ndiagnostic criteria, the model provides a useful tool to uncover a patient's\nlatent sepsis trajectory and to identify high-risk patients in whom more\naggressive therapy may be indicated.\n",
"title": "Modeling sepsis progression using hidden Markov models"
}
| null | null |
[
"Statistics",
"Quantitative Biology"
] | null | true | null |
10909
| null |
Validated
| null | null |
null |
{
"abstract": " Experimentalists have observed phenotypic variability in isogenic bacteria\npopulations. We explore the hypothesis that in fluctuating environments this\nvariability is tuned to maximize a bacterium's expected log growth rate,\npotentially aided by epigenetic markers that store information about past\nenvironments. We show that, in a complex, memoryful environment, the maximal\nexpected log growth rate is linear in the instantaneous predictive\ninformation---the mutual information between a bacterium's epigenetic markers\nand future environmental states. Hence, under resource constraints, optimal\nepigenetic markers are causal states---the minimal sufficient statistics for\nprediction. This is the minimal amount of information about the past needed to\npredict the future as well as possible. We suggest new theoretical\ninvestigations into and new experiments on bacteria phenotypic bet-hedging in\nfluctuating complex environments.\n",
"title": "Optimized Bacteria are Environmental Prediction Engines"
}
| null | null | null | null | true | null |
10910
| null |
Default
| null | null |
null |
{
"abstract": " Style transfer methods have achieved significant success in recent years with\nthe use of convolutional neural networks. However, many of these methods\nconcentrate on artistic style transfer with few constraints on the output image\nappearance. We address the challenging problem of transferring face texture\nfrom a style face image to a content face image in a photorealistic manner\nwithout changing the identity of the original content image. Our framework for\nface texture transfer (FaceTex) augments the prior work of MRF-CNN with a novel\nfacial semantic regularization that incorporates a face prior regularization\nsmoothly suppressing the changes around facial meso-structures (e.g eyes, nose\nand mouth) and a facial structure loss function which implicitly preserves the\nfacial structure so that face texture can be transferred without changing the\noriginal identity. We demonstrate results on face images and compare our\napproach with recent state-of-the-art methods. Our results demonstrate superior\ntexture transfer because of the ability to maintain the identity of the\noriginal face image.\n",
"title": "Photo-realistic Facial Texture Transfer"
}
| null | null |
[
"Computer Science"
] | null | true | null |
10911
| null |
Validated
| null | null |
null |
{
"abstract": " This paper proposes an innovative method for segmentation of skin lesions in\ndermoscopy images developed by the authors, based on fuzzy classification of\npixels and histogram thresholding.\n",
"title": "Segmentation of skin lesions based on fuzzy classification of pixels and histogram thresholding"
}
| null | null | null | null | true | null |
10912
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we explore deep reinforcement learning algorithms for\nvision-based robotic grasping. Model-free deep reinforcement learning (RL) has\nbeen successfully applied to a range of challenging environments, but the\nproliferation of algorithms makes it difficult to discern which particular\napproach would be best suited for a rich, diverse task like grasping. To answer\nthis question, we propose a simulated benchmark for robotic grasping that\nemphasizes off-policy learning and generalization to unseen objects. Off-policy\nlearning enables utilization of grasping data over a wide variety of objects,\nand diversity is important to enable the method to generalize to new objects\nthat were not seen during training. We evaluate the benchmark tasks against a\nvariety of Q-function estimation methods, a method previously proposed for\nrobotic grasping with deep neural network models, and a novel approach based on\na combination of Monte Carlo return estimation and an off-policy correction.\nOur results indicate that several simple methods provide a surprisingly strong\ncompetitor to popular algorithms such as double Q-learning, and our analysis of\nstability sheds light on the relative tradeoffs between the algorithms.\n",
"title": "Deep Reinforcement Learning for Vision-Based Robotic Grasping: A Simulated Comparative Evaluation of Off-Policy Methods"
}
| null | null | null | null | true | null |
10913
| null |
Default
| null | null |
null |
{
"abstract": " The key issues pertaining to collection of epidemic disease data for our\nanalysis purposes are that it is a labour intensive, time consuming and\nexpensive process resulting in availability of sparse sample data which we use\nto develop prediction models. To address this sparse data issue, we present\nnovel Incremental Transductive methods to circumvent the data collection\nprocess by applying previously acquired data to provide consistent,\nconfidence-based labelling alternatives to field survey research. We\ninvestigated various reasoning approaches for semisupervised machine learning\nincluding Bayesian models for labelling data. The results show that using the\nproposed methods, we can label instances of data with a class of vector density\nat a high level of confidence. By applying the Liberal and Strict Training\nApproaches, we provide a labelling and classification alternative to standalone\nalgorithms. The methods in this paper are components in the process of reducing\nthe proliferation of the Schistosomiasis disease and its effects.\n",
"title": "Incremental Transductive Learning Approaches to Schistosomiasis Vector Classification"
}
| null | null |
[
"Computer Science"
] | null | true | null |
10914
| null |
Validated
| null | null |
null |
{
"abstract": " Molecular beam epitaxy technique has been used to deposit a single layer and\na bilayer of MoSe 2 on sapphire. Extensive characterizations including in-situ\nand ex-situ measurements show that the layered MoSe 2 grows in a scalable\nmanner on the substrate and reveals characteristics of a stoichiometric\n2H-phase. The layered MoSe 2 exhibits polycrystalline features with domains\nseparated by defects and boundaries. Temperature and magnetic field dependent\nresistivity measurements unveil a carrier hopping character described within\ntwo-dimensional variable range hopping mechanism. Moreover, a negative\nmagnetoresistance was observed, stressing a fascinating feature of the charge\ntransport under the application of a magnetic field in the layered MoSe 2\nsystem. This negative magnetoresistance observed at millimeter-scale is similar\nto that observed recently at room temperature inWS2 flakes at a micrometer\nscale [Zhang et al., Appl. Phys. Lett. 108, 153114 (2016)]. This scalability\nhighlights the fact that the underlying physical mechanism is intrinsic to\nthese two-dimensional materials and occurs at very short scale.\n",
"title": "Millimeter-scale layered MoSe2 grown on sapphire and evidence for negative magnetoresistance"
}
| null | null | null | null | true | null |
10915
| null |
Default
| null | null |
null |
{
"abstract": " We investigate an end-to-end method for automatically inducing task-based\ndialogue systems from small amounts of unannotated dialogue data. It combines\nan incremental semantic grammar - Dynamic Syntax and Type Theory with Records\n(DS-TTR) - with Reinforcement Learning (RL), where language generation and\ndialogue management are a joint decision problem. The systems thus produced are\nincremental: dialogues are processed word-by-word, shown previously to be\nessential in supporting natural, spontaneous dialogue. We hypothesised that the\nrich linguistic knowledge within the grammar should enable a combinatorially\nlarge number of dialogue variations to be processed, even when trained on very\nfew dialogues. Our experiments show that our model can process 74% of the\nFacebook AI bAbI dataset even when trained on only 0.13% of the data (5\ndialogues). It can in addition process 65% of bAbI+, a corpus we created by\nsystematically adding incremental dialogue phenomena such as restarts and\nself-corrections to bAbI. We compare our model with a state-of-the-art\nretrieval model, MemN2N. We find that, in terms of semantic accuracy, MemN2N\nshows very poor robustness to the bAbI+ transformations even when trained on\nthe full bAbI dataset.\n",
"title": "Bootstrapping incremental dialogue systems from minimal data: the generalisation power of dialogue grammars"
}
| null | null | null | null | true | null |
10916
| null |
Default
| null | null |
null |
{
"abstract": " Multi-label classification is a practical yet challenging task in machine\nlearning related fields, since it requires the prediction of more than one\nlabel category for each input instance. We propose a novel deep neural networks\n(DNN) based model, Canonical Correlated AutoEncoder (C2AE), for solving this\ntask. Aiming at better relating feature and label domain data for improved\nclassification, we uniquely perform joint feature and label embedding by\nderiving a deep latent space, followed by the introduction of label-correlation\nsensitive loss function for recovering the predicted label outputs. Our C2AE is\nachieved by integrating the DNN architectures of canonical correlation analysis\nand autoencoder, which allows end-to-end learning and prediction with the\nability to exploit label dependency. Moreover, our C2AE can be easily extended\nto address the learning problem with missing labels. Our experiments on\nmultiple datasets with different scales confirm the effectiveness and\nrobustness of our proposed method, which is shown to perform favorably against\nstate-of-the-art methods for multi-label classification.\n",
"title": "Learning Deep Latent Spaces for Multi-Label Classification"
}
| null | null | null | null | true | null |
10917
| null |
Default
| null | null |
null |
{
"abstract": " The main goal of this study is to extract a set of brain networks in multiple\ntime-resolutions to analyze the connectivity patterns among the anatomic\nregions for a given cognitive task. We suggest a deep architecture which learns\nthe natural groupings of the connectivity patterns of human brain in multiple\ntime-resolutions. The suggested architecture is tested on task data set of\nHuman Connectome Project (HCP) where we extract multi-resolution networks, each\nof which corresponds to a cognitive task. At the first level of this\narchitecture, we decompose the fMRI signal into multiple sub-bands using\nwavelet decompositions. At the second level, for each sub-band, we estimate a\nbrain network extracted from short time windows of the fMRI signal. At the\nthird level, we feed the adjacency matrices of each mesh network at each\ntime-resolution into an unsupervised deep learning algorithm, namely, a Stacked\nDe- noising Auto-Encoder (SDAE). The outputs of the SDAE provide a compact\nconnectivity representation for each time window at each sub-band of the fMRI\nsignal. We concatenate the learned representations of all sub-bands at each\nwindow and cluster them by a hierarchical algorithm to find the natural\ngroupings among the windows. We observe that each cluster represents a\ncognitive task with a performance of 93% Rand Index and 71% Adjusted Rand\nIndex. We visualize the mean values and the precisions of the networks at each\ncomponent of the cluster mixture. The mean brain networks at cluster centers\nshow the variations among cognitive tasks and the precision of each cluster\nshows the within cluster variability of networks, across the subjects.\n",
"title": "Encoding Multi-Resolution Brain Networks Using Unsupervised Deep Learning"
}
| null | null | null | null | true | null |
10918
| null |
Default
| null | null |
null |
{
"abstract": " Combinatorial filters have been the subject of increasing interest from the\nrobotics community in recent years. This paper considers automatic reduction of\ncombinatorial filters to a given size, even if that reduction necessitates\nchanges to the filter's behavior. We introduce an algorithmic problem called\nimproper filter reduction, in which the input is a combinatorial filter F along\nwith an integer k representing the target size. The output is another\ncombinatorial filter F' with at most k states, such that the difference in\nbehavior between F and F' is minimal. We present two metrics for measuring the\ndistance between pairs of filters, describe dynamic programming algorithms for\ncomputing these distances, and show that improper filter reduction is NP-hard\nunder these metrics. We then describe two heuristic algorithms for improper\nfilter reduction, one greedy sequential approach, and one randomized global\napproach based on prior work on weighted improper graph coloring. We have\nimplemented these algorithms and analyze the results of three sets of\nexperiments.\n",
"title": "Improper Filter Reduction"
}
| null | null |
[
"Computer Science"
] | null | true | null |
10919
| null |
Validated
| null | null |
null |
{
"abstract": " The collective behavior of active semiflexible filaments is studied with a\nmodel of tangentially driven self-propelled worm-like chains. The combination\nof excluded-volume interactions and self-propulsion leads to several distinct\ndynamic phases as a function of bending rigidity, activity, and aspect ratio of\nindividual filaments. We consider first the case of intermediate filament\ndensity. For high-aspect-ratio filaments, we identify a transition with\nincreasing propulsion from a state of free-swimming filaments to a state of\nspiraled filaments with nearly frozen translational motion. For lower aspect\nratios, this gas-of-spirals phase is suppressed with growing density due to\nfilament collisions; instead, filaments form clusters similar to self-propelled\nrods, as activity increases. Finite bending rigidity strongly effects the\ndynamics and phase behavior. Flexible filaments form small and transient\nclusters, while stiffer filaments organize into giant clusters, similarly as\nself-propelled rods, but with a reentrant phase behavior from giant to smaller\nclusters as activity becomes large enough to bend the filaments. For high\nfilament densities, we identify a nearly frozen jamming state at low\nactivities, a nematic laning state at intermediate activities, and an\nactive-turbulence state at high activities. The latter state is characterized\nby a power-law decay of the energy spectrum as a function of wave number. The\nresulting phase diagrams encapsulate tunable non-equilibrium steady states that\ncan be used in the organization of living matter.\n",
"title": "Collective Dynamics of Self-propelled Semiflexible Filaments"
}
| null | null | null | null | true | null |
10920
| null |
Default
| null | null |
null |
{
"abstract": " We compare the following two sources of poor coverage of post-model-selection\nconfidence intervals: the preliminary data-based model selection sometimes\nchooses the wrong model and the data used to choose the model is re-used for\nthe construction of the confidence interval.\n",
"title": "Two sources of poor coverage of confidence intervals after model selection"
}
| null | null | null | null | true | null |
10921
| null |
Default
| null | null |
null |
{
"abstract": " Networked data, in which every training example involves two objects and may\nshare some common objects with others, is used in many machine learning tasks\nsuch as learning to rank and link prediction. A challenge of learning from\nnetworked examples is that target values are not known for some pairs of\nobjects. In this case, neither the classical i.i.d.\\ assumption nor techniques\nbased on complete U-statistics can be used. Most existing theoretical results\nof this problem only deal with the classical empirical risk minimization (ERM)\nprinciple that always weights every example equally, but this strategy leads to\nunsatisfactory bounds. We consider general weighted ERM and show new universal\nrisk bounds for this problem. These new bounds naturally define an optimization\nproblem which leads to appropriate weights for networked examples. Though this\noptimization problem is not convex in general, we devise a new fully\npolynomial-time approximation scheme (FPTAS) to solve it.\n",
"title": "On the ERM Principle with Networked Data"
}
| null | null | null | null | true | null |
10922
| null |
Default
| null | null |
null |
{
"abstract": " Spatially extended systems can support local transient excitations in which\njust a part of the system is excited. The mechanisms reported so far are local\nexcitability and excitation of a localized structure. Here we introduce an\nalternative mechanism based on the coexistence of two homogeneous stable states\nand spatial coupling. We show the existence of a threshold for perturbations of\nthe homogeneous state. Sub-threshold perturbations decay exponentially.\nSuper-threshold perturbations induce the emergence of a long-lived structure\nformed by two back to back fronts that join the two homogeneous states. While\nin typical excitability the trajectory follows the remnants of a limit cycle,\nhere reinjection is provided by front interaction, such that fronts slowly\napproach each other until eventually annihilating. This front-mediated\nmechanism shows that extended systems with no oscillatory regimes can display\nexcitability.\n",
"title": "Front interaction induces excitable behavior"
}
| null | null | null | null | true | null |
10923
| null |
Default
| null | null |
null |
{
"abstract": " Carbon nanotubes are modeled as point configurations and investigated by\nminimizing configurational energies including two-and three-body interactions.\nOptimal configurations are identified with local minima and their fine geometry\nis fully characterized in terms of lower-dimensional problems. Under moderate\ntension, we prove the existence of periodic local minimizers, which indeed\nvalidates the so-called Cauchy-Born rule in this setting.\n",
"title": "Characterization of optimal carbon nanotubes under stretching and validation of the Cauchy-Born rule"
}
| null | null | null | null | true | null |
10924
| null |
Default
| null | null |
null |
{
"abstract": " We investigate the similarities of pairs of articles which are co-cited at\nthe different co-citation levels of the journal, article, section, paragraph,\nsentence and bracket. Our results indicate that textual similarity,\nintellectual overlap (shared references), author overlap (shared authors),\nproximity in publication time all rise monotonically as the co-citation level\ngets lower (from journal to bracket). While the main gain in similarity happens\nwhen moving from journal to article co-citation, all level changes entail an\nincrease in similarity, especially section to paragraph and paragraph to\nsentence/bracket levels. We compare results from four journals over the years\n2010-2015: Cell, the European Journal of Operational Research, Physics Letters\nB and Research Policy, with consistent general outcomes and some interesting\ndifferences. Our findings motivate the use of granular co-citation information\nas defined by meaningful units of text, with implications for, among others,\nthe elaboration of maps of science and the retrieval of scholarly literature.\n",
"title": "The Closer the Better: Similarity of Publication Pairs at Different Co-Citation Levels"
}
| null | null |
[
"Computer Science"
] | null | true | null |
10925
| null |
Validated
| null | null |
null |
{
"abstract": " Complementary auxiliary basis sets for F12 explicitly correlated calculations\nappear to be more transferable between orbital basis sets than has been\ngenerally assumed. We also find that aVnZ-F12 basis sets, originally developed\nwith anionic systems in mind, appear to be superior for noncovalent\ninteractions as well, and propose a suitable CABS sequence for them.\n",
"title": "MP2-F12 Basis Set Convergence for the S66 Noncovalent Interactions Benchmark: Transferability of the Complementary Auxiliary Basis Set (CABS)"
}
| null | null | null | null | true | null |
10926
| null |
Default
| null | null |
null |
{
"abstract": " Comparing with traditional learning criteria, such as mean square error\n(MSE), the minimum error entropy (MEE) criterion is superior in nonlinear and\nnon-Gaussian signal processing and machine learning. The argument of the\nlogarithm in Renyis entropy estimator, called information potential (IP), is a\npopular MEE cost in information theoretic learning (ITL). The computational\ncomplexity of IP is however quadratic in terms of sample number due to double\nsummation. This creates computational bottlenecks especially for large-scale\ndatasets. To address this problem, in this work we propose an efficient\nquantization approach to reduce the computational burden of IP, which decreases\nthe complexity from O(N*N) to O (MN) with M << N. The new learning criterion is\ncalled the quantized MEE (QMEE). Some basic properties of QMEE are presented.\nIllustrative examples are provided to verify the excellent performance of QMEE.\n",
"title": "Quantized Minimum Error Entropy Criterion"
}
| null | null | null | null | true | null |
10927
| null |
Default
| null | null |
null |
{
"abstract": " A novel predictor for traffic flow forecasting, namely spatio-temporal\nBayesian network predictor, is proposed. Unlike existing methods, our approach\nincorporates all the spatial and temporal information available in a\ntransportation network to carry our traffic flow forecasting of the current\nsite. The Pearson correlation coefficient is adopted to rank the input\nvariables (traffic flows) for prediction, and the best-first strategy is\nemployed to select a subset as the cause nodes of a Bayesian network. Given the\nderived cause nodes and the corresponding effect node in the spatio-temporal\nBayesian network, a Gaussian Mixture Model is applied to describe the\nstatistical relationship between the input and output. Finally, traffic flow\nforecasting is performed under the criterion of Minimum Mean Square Error\n(M.M.S.E.). Experimental results with the urban vehicular flow data of Beijing\ndemonstrate the effectiveness of our presented spatio-temporal Bayesian network\npredictor.\n",
"title": "Traffic Flow Forecasting Using a Spatio-Temporal Bayesian Network Predictor"
}
| null | null |
[
"Computer Science",
"Statistics"
] | null | true | null |
10928
| null |
Validated
| null | null |
null |
{
"abstract": " A person dependent network, called an AlterEgo net, is proposed for\ndevelopment. The networks are created per person. It receives at input an\nobject descriptions and outputs a simulation of the internal person's\nrepresentation of the objects. The network generates a textual stream\nresembling the narrative stream of consciousness depicting multitudinous\nthoughts and feelings related to a perceived object. In this way, the object is\ndescribed not by a 'static' set of its properties, like a dictionary, but by\nthe stream of words and word combinations referring to the object. The network\nsimulates a person's dialogue with a representation of the object. It is based\non an introduced algorithmic scheme, where perception is modeled by two\ninteracting iterative cycles, reminding one respectively the forward and\nbackward propagation executed at training convolution neural networks. The\n'forward' iterations generate a stream representing the 'internal world' of a\nhuman. The 'backward' iterations generate a stream representing an internal\nrepresentation of the object. People perceive the world differently. Tuning\nAlterEgo nets to a specific person or group of persons, will allow simulation\nof their thoughts and feelings. Thereby these nets is potentially a new human\naugmentation technology for various applications.\n",
"title": "AlteregoNets: a way to human augmentation"
}
| null | null | null | null | true | null |
10929
| null |
Default
| null | null |
null |
{
"abstract": " Univalent homotopy type theory (HoTT) may be seen as a language for the\ncategory of $\\infty$-groupoids. It is being developed as a new foundation for\nmathematics and as an internal language for (elementary) higher toposes. We\ndevelop the theory of factorization systems, reflective subuniverses, and\nmodalities in homotopy type theory, including their construction using a\n\"localization\" higher inductive type. This produces in particular the\n($n$-connected, $n$-truncated) factorization system as well as internal\npresentations of subtoposes, through lex modalities. We also develop the\nsemantics of these constructions.\n",
"title": "Modalities in homotopy type theory"
}
| null | null | null | null | true | null |
10930
| null |
Default
| null | null |
null |
{
"abstract": " We analyze the dynamics of periodically-driven (Floquet) Hamiltonians with\nshort- and long-range interactions, finding clear evidence for a thermalization\ntime, $\\tau^*$, that increases exponentially with the drive frequency. We\nobserve this behavior, both in systems with short-ranged interactions, where\nour results are consistent with rigorous bounds, and in systems with long-range\ninteractions, where such bounds do not exist at present. Using a combination of\nheating and entanglement dynamics, we explicitly extract the effective energy\nscale controlling the rate of thermalization. Finally, we demonstrate that for\ntimes shorter than $\\tau^*$, the dynamics of the system is well-approximated by\nevolution under a time-independent Hamiltonian $D_{\\mathrm{eff}}$, for both\nshort- and long-range interacting systems.\n",
"title": "Exponentially Slow Heating in Short and Long-range Interacting Floquet Systems"
}
| null | null | null | null | true | null |
10931
| null |
Default
| null | null |
null |
{
"abstract": " For $a/q\\in\\mathbb{Q}$ the Estermann function is defined as\n$D(s,a/q):=\\sum_{n\\geq1}d(n)n^{-s}\\operatorname{e}(n\\frac aq)$ if $\\Re(s)>1$\nand by meromorphic continuation otherwise. For $q$ prime, we compute the\nmoments of $D(s,a/q)$ at the central point $s=1/2$, when averaging over $1\\leq\na<q$.\nAs a consequence we deduce the asymptotic for the iterated moment of\nDirichlet $L$-functions $\\sum_{\\chi_1,\\dots,\\chi_k\\mod\nq}|L(\\frac12,\\chi_1)|^2\\cdots |L(\\frac12,\\chi_k)|^2|L(\\frac12,\\chi_1\\cdots\n\\chi_k)|^2$, obtaining a power saving error term.\nAlso, we compute the moments of certain functions defined in terms of\ncontinued fractions. For example, writing $f_{\\pm}(a/q):=\\sum_{j=0}^r\n(\\pm1)^jb_j$ where $[0;b_0,\\dots,b_r]$ is the continued fraction expansion of\n$a/q$ we prove that for $k\\geq2$ and $q$ primes one has\n$\\sum_{a=1}^{q-1}f_{\\pm}(a/q)^k\\sim2 \\frac{\\zeta(k)^2}{\\zeta(2k)} q^k$ as\n$q\\to\\infty$.\n",
"title": "High moments of the Estermann function"
}
| null | null | null | null | true | null |
10932
| null |
Default
| null | null |
null |
{
"abstract": " The game-theoretic risk management framework put forth in the precursor\nreports \"Towards a Theory of Games with Payoffs that are\nProbability-Distributions\" (arXiv:1506.07368 [q-fin.EC]) and \"Algorithms to\nCompute Nash-Equilibria in Games with Distributions as Payoffs\"\n(arXiv:1511.08591v1 [q-fin.EC]) is herein concluded by discussing how to\nintegrate the previously developed theory into risk management processes. To\nthis end, we discuss how loss models (primarily but not exclusively\nnon-parametric) can be constructed from data. Furthermore, hints are given on\nhow a meaningful game theoretic model can be set up, and how it can be used in\nvarious stages of the ISO 27000 risk management process. Examples related to\nadvanced persistent threats and social engineering are given. We conclude by a\ndiscussion on the meaning and practical use of (mixed) Nash equilibria\nequilibria for risk management.\n",
"title": "On Game-Theoretic Risk Management (Part Three) - Modeling and Applications"
}
| null | null |
[
"Mathematics",
"Statistics"
] | null | true | null |
10933
| null |
Validated
| null | null |
null |
{
"abstract": " We present results from a multiwavelength study of the blazar PKS 1954-388 at\nradio, UV, X-ray, and gamma-ray energies. A RadioAstron observation at 1.66 GHz\nin June 2012 resulted in the detection of interferometric fringes on baselines\nof 6.2 Earth-diameters. This suggests a source frame brightness temperature of\ngreater than 2x10^12 K, well in excess of both equipartition and inverse\nCompton limits and implying the existence of Doppler boosting in the core. An\n8.4 GHz TANAMI VLBI image, made less than a month after the RadioAstron\nobservations, is consistent with a previously reported superluminal motion for\na jet component. Flux density monitoring with the Australia Telescope Compact\nArray confirms previous evidence for long-term variability that increases with\nobserving frequency. A search for more rapid variability revealed no evidence\nfor significant day-scale flux density variation. The ATCA light-curve reveals\na strong radio flare beginning in late 2013 which peaks higher, and earlier, at\nhigher frequencies. Comparison with the Fermi gamma-ray light-curve indicates\nthis followed ~9 months after the start of a prolonged gamma-ray high-state --\na radio lag comparable to that seen in other blazars. The multiwavelength data\nare combined to derive a Spectral Energy Distribution, which is fitted by a\none-zone synchrotron-self-Compton (SSC) model with the addition of external\nCompton (EC) emission.\n",
"title": "PKS 1954-388: RadioAstron Detection on 80,000 km Baselines and Multiwavelength Observations"
}
| null | null | null | null | true | null |
10934
| null |
Default
| null | null |
null |
{
"abstract": " One of the major hurdles toward automatic semantic understanding of computer\nprograms is the lack of knowledge about what constitutes functional equivalence\nof code segments. We postulate that a sound knowledgebase can be used to\ndeductively understand code segments in a hierarchical fashion by first\nde-constructing a code and then reconstructing it from elementary knowledge and\nequivalence rules of elementary code segments. The approach can also be\nengineered to produce computable programs from conceptual and abstract\nalgorithms as an inverse function. In this paper, we introduce the core idea\nbehind the MindReader online assessment system that is able to understand a\nwide variety of elementary algorithms students learn in their entry level\nprogramming classes such as Java, C++ and Python. The MindReader system is able\nto assess student assignments and guide them how to develop correct and better\ncode in real time without human assistance.\n",
"title": "Smart Assessment of and Tutoring for Computational Thinking MOOC Assignments using MindReader"
}
| null | null | null | null | true | null |
10935
| null |
Default
| null | null |
null |
{
"abstract": " In this paper we describe EasyInterface, an open-source toolkit for rapid\ndevelopment of web-based graphical user interfaces (GUIs). This toolkit\naddresses the need of researchers to make their research prototype tools\navailable to the community, and integrating them in a common environment,\nrapidly and without being familiar with web programming or GUI libraries in\ngeneral. If a tool can be executed from a command-line and its output goes to\nthe standard output, then in few minutes one can make it accessible via a\nweb-interface or within Eclipse. Moreover, the toolkit defines a text-based\nlanguage that can be used to get more sophisticated GUIs, e.g., syntax\nhighlighting, dialog boxes, user interactions, etc. EasyInterface was\noriginally developed for building a common frontend for tools developed in the\nEnvisage project.\n",
"title": "EasyInterface: A toolkit for rapid development of GUIs for research prototype tools"
}
| null | null | null | null | true | null |
10936
| null |
Default
| null | null |
null |
{
"abstract": " We introduce the notion of the essential tangent bundle of a parametrized\nmeasure model and the notion of reduced Fisher metric on a (possibly singular)\n2-integrable measure model. Using these notions and a new characterization of\n$k$-integrable parametrized measure models, we extend the Cramér-Rao\ninequality to $2$-integrable (possibly singular) statistical models for general\n$\\varphi$-estimations, where $\\varphi$ is a $V$-valued feature function and $V$\nis a topological vector space. Thus we derive an intrinsic Cramér-Rao\ninequality in the most general terms of parametric statistics.\n",
"title": "The Cramér-Rao inequality on singular statistical models I"
}
| null | null | null | null | true | null |
10937
| null |
Default
| null | null |
null |
{
"abstract": " Machine learning applications often require hyperparameter tuning. The\nhyperparameters usually drive both the efficiency of the model training process\nand the resulting model quality. For hyperparameter tuning, machine learning\nalgorithms are complex black-boxes. This creates a class of challenging\noptimization problems, whose objective functions tend to be nonsmooth,\ndiscontinuous, unpredictably varying in computational expense, and include\ncontinuous, categorical, and/or integer variables. Further, function\nevaluations can fail for a variety of reasons including numerical difficulties\nor hardware failures. Additionally, not all hyperparameter value combinations\nare compatible, which creates so called hidden constraints. Robust and\nefficient optimization algorithms are needed for hyperparameter tuning. In this\npaper we present an automated parallel derivative-free optimization framework\ncalled \\textbf{Autotune}, which combines a number of specialized sampling and\nsearch methods that are very effective in tuning machine learning models\ndespite these challenges. Autotune provides significantly improved models over\nusing default hyperparameter settings with minimal user interaction on\nreal-world applications. Given the inherent expense of training numerous\ncandidate models, we demonstrate the effectiveness of Autotune's search methods\nand the efficient distributed and parallel paradigms for training and tuning\nmodels, and also discuss the resource trade-offs associated with the ability to\nboth distribute the training process and parallelize the tuning process.\n",
"title": "Autotune: A Derivative-free Optimization Framework for Hyperparameter Tuning"
}
| null | null | null | null | true | null |
10938
| null |
Default
| null | null |
null |
{
"abstract": " We finish the classification, begun in two earlier papers, of all simple\nfusion systems over finite nonabelian $p$-groups with an abelian subgroup of\nindex $p$. In particular, this gives many new examples illustrating the\nenormous variety of exotic examples that can arise. In addition, we classify\nall simple fusion systems over infinite nonabelian discrete $p$-toral groups\nwith an abelian subgroup of index $p$. In all of these cases (finite or\ninfinite), we reduce the problem to one of listing all $\\mathbb{F}_pG$-modules\n(for $G$ finite) satisfying certain conditions: a problem which was solved in\nthe earlier paper by Craven, Oliver, and Semeraro using the classification of\nfinite simple groups.\n",
"title": "Reduced fusion systems over $p$-groups with abelian subgroup of index $p$: III"
}
| null | null | null | null | true | null |
10939
| null |
Default
| null | null |
null |
{
"abstract": " Establishing accurate morphological measurements of galaxies in a reasonable\namount of time for future big-data surveys such as EUCLID, the Large Synoptic\nSurvey Telescope or the Wide Field Infrared Survey Telescope is a challenge.\nBecause of its high level of abstraction with little human intervention, deep\nlearning appears to be a promising approach. Deep learning is a rapidly growing\ndiscipline that models high-level patterns in data as complex multilayered\nnetworks. In this work we test the ability of deep convolutional networks to\nprovide parametric properties of Hubble Space Telescope like galaxies\n(half-light radii, Sersic indices, total flux etc..). We simulate a set of\ngalaxies including point spread function and realistic noise from the CANDELS\nsurvey and try to recover the main galaxy parameters using deep-learning. We\ncom- pare the results with the ones obtained with the commonly used profile\nfitting based software GALFIT. This way showing that with our method we obtain\nresults at least equally good as the ones obtained with GALFIT but, once\ntrained, with a factor 5 hundred time faster.\n",
"title": "Deep learning for studies of galaxy morphology"
}
| null | null | null | null | true | null |
10940
| null |
Default
| null | null |
null |
{
"abstract": " Multi-start algorithms are a common and effective tool for metaheuristic\nsearches. In this paper we amplify multi-start capabilities by employing the\nparallel processing power of the graphics processer unit (GPU) to quickly\ngenerate a diverse starting set of solutions for the Unconstrained Binary\nQuadratic Optimization Problem which are evaluated and used to implement\nscreening methods to select solutions for further optimization. This method is\nimplemented as an initial high quality solution generation phase prior to a\nsecondary steepest ascent search and a comparison of results to best known\napproaches on benchmark unconstrained binary quadratic problems demonstrates\nthat GPU-enabled diversified multi-start with screening quickly yields very\ngood results.\n",
"title": "A Diversified Multi-Start Algorithm for Unconstrained Binary Quadratic Problems Leveraging the Graphics Processor Unit"
}
| null | null | null | null | true | null |
10941
| null |
Default
| null | null |
null |
{
"abstract": " Estimation of the number of endmembers existing in a scene constitutes a\ncritical task in the hyperspectral unmixing process. The accuracy of this\nestimate plays a crucial role in subsequent unsupervised unmixing steps i.e.,\nthe derivation of the spectral signatures of the endmembers (endmembers'\nextraction) and the estimation of the abundance fractions of the pixels. A\ncommon practice amply followed in literature is to treat endmembers' number\nestimation and unmixing, independently as two separate tasks, providing the\noutcome of the former as input to the latter. In this paper, we go beyond this\ncomputationally demanding strategy. More precisely, we set forth a multiple\nconstrained optimization framework, which encapsulates endmembers' number\nestimation and unsupervised unmixing in a single task. This is attained by\nsuitably formulating the problem via a low-rank and sparse nonnegative matrix\nfactorization rationale, where low-rankness is promoted with the use of a\nsophisticated $\\ell_2/\\ell_1$ norm penalty term. An alternating proximal\nalgorithm is then proposed for minimizing the emerging cost function. The\nresults obtained by simulated and real data experiments verify the\neffectiveness of the proposed approach.\n",
"title": "Low-rank and Sparse NMF for Joint Endmembers' Number Estimation and Blind Unmixing of Hyperspectral Images"
}
| null | null | null | null | true | null |
10942
| null |
Default
| null | null |
null |
{
"abstract": " The problem of three-user multiple-access channel (MAC) with noiseless\nfeedback is investigated. A new coding strategy is presented. The coding scheme\nbuilds upon the natural extension of the Cover-Leung (CL) scheme; and uses\nquasi-linear codes. A new single-letter achievable rate region is derived. The\nnew achievable region strictly contains the CL region. This is shown through an\nexample. In this example, the coding scheme achieves optimality in terms of\ntransmission rates. It is shown that any optimality achieving scheme for this\nexample must have a specific algebraic structure. Particularly, the codebooks\nmust be closed under binary addition.\n",
"title": "On the Necessity of Structured Codes for Communications over MAC with Feedback"
}
| null | null |
[
"Computer Science"
] | null | true | null |
10943
| null |
Validated
| null | null |
null |
{
"abstract": " In the last two decades recurrence plots (RPs) were introduced in many\ndifferent scientific disciplines. It turned out how powerful this method is.\nAfter introducing approaches of quantification of RPs and by the study of\nrelationships between RPs and fundamental properties of dynamical systems, this\nmethod attracted even more attention. After 20 years of RPs it is time to\nsummarise this development in a historical context.\n",
"title": "Historical Review of Recurrence Plots"
}
| null | null | null | null | true | null |
10944
| null |
Default
| null | null |
null |
{
"abstract": " The meta distribution of the signal-to-interference ratio (SIR) provides\nfine-grained information about the performance of individual links in a\nwireless network. This paper focuses on the analysis of the meta distribution\nof the SIR for both the cellular network uplink and downlink with fractional\npower control. For the uplink scenario, an approximation of the interfering\nuser point process with a non-homogeneous Poisson point process is used. The\nmoments of the meta distribution for both scenarios are calculated. Some\nbounds, the analytical expression, the mean local delay, and the beta\napproximation of the meta distribution are provided. The results give\ninteresting insights into the effect of the power control in both the uplink\nand downlink. Detailed simulations show that the approximations made in the\nanalysis are well justified.\n",
"title": "The Meta Distribution of the SIR for Cellular Networks with Power Control"
}
| null | null |
[
"Computer Science"
] | null | true | null |
10945
| null |
Validated
| null | null |
null |
{
"abstract": " The transport characteristics across the pulsed laser deposited\nNi0.07Zn0.93O/Mg0.21Zn0.79O heterojunction exhibits p-n type semiconducting\nproperties at 10 K while at 100 K, its characteristics become similar to that\nof an n-n junction. The reason for the same is attributed to the role of larger\nelectronegativity of Ni as compared to Mg at 10 K and ionization of impurity\nstates at 100 K. The above behavior is confirmed by performing the Hall\nmeasurements.\n",
"title": "Temperature induced transition from p-n to n-n electronic behavior in Ni0.07Zn0.93O/Mg0.21Zn0.79O heterojunction"
}
| null | null | null | null | true | null |
10946
| null |
Default
| null | null |
null |
{
"abstract": " Large redshift surveys of galaxies and clusters are providing the first\nopportunities to search for distortions in the observed pattern of large-scale\nstructure due to such effects as gravitational redshift. We focus on non-linear\nscales and apply a quasi-Newtonian approach using N-body simulations to predict\nthe small asymmetries in the cross-correlation function of two galaxy different\npopulations. Following recent work by Bonvin et al., Zhao and Peacock and\nKaiser on galaxy clusters, we include effects which enter at the same order as\ngravitational redshift: the transverse Doppler effect, light-cone effects,\nrelativistic beaming, luminosity distance perturbation and wide-angle effects.\nWe find that all these effects cause asymmetries in the cross-correlation\nfunctions. Quantifying these asymmetries, we find that the total effect is\ndominated by the gravitational redshift and luminosity distance perturbation at\nsmall and large scales, respectively. By adding additional subresolution\nmodelling of galaxy structure to the large-scale structure information, we find\nthat the signal is significantly increased, indicating that structure on the\nsmallest scales is important and should be included. We report on comparison of\nour simulation results with measurements from the SDSS/BOSS galaxy redshift\nsurvey in a companion paper.\n",
"title": "N-body simulations of gravitational redshifts and other relativistic distortions of galaxy clustering"
}
| null | null | null | null | true | null |
10947
| null |
Default
| null | null |
null |
{
"abstract": " Learning high quality class representations from few examples is a key\nproblem in metric-learning approaches to few-shot learning. To accomplish this,\nwe introduce a novel architecture where class representations are conditioned\nfor each few-shot trial based on a target image. We also deviate from\ntraditional metric-learning approaches by training a network to perform\ncomparisons between classes rather than relying on a static metric comparison.\nThis allows the network to decide what aspects of each class are important for\nthe comparison at hand. We find that this flexible architecture works well in\npractice, achieving state-of-the-art performance on the Caltech-UCSD birds\nfine-grained classification task.\n",
"title": "Few-Shot Learning with Metric-Agnostic Conditional Embeddings"
}
| null | null | null | null | true | null |
10948
| null |
Default
| null | null |
null |
{
"abstract": " Gradient coding is a technique for straggler mitigation in distributed\nlearning. In this paper we design novel gradient codes using tools from\nclassical coding theory, namely, cyclic MDS codes, which compare favourably\nwith existing solutions, both in the applicable range of parameters and in the\ncomplexity of the involved algorithms. Second, we introduce an approximate\nvariant of the gradient coding problem, in which we settle for approximate\ngradient computation instead of the exact one. This approach enables graceful\ndegradation, i.e., the $\\ell_2$ error of the approximate gradient is a\ndecreasing function of the number of stragglers. Our main result is that the\nnormalized adjacency matrix of an expander graph can yield excellent\napproximate gradient codes, and that this approach allows us to perform\nsignificantly less computation compared to exact gradient coding. We\nexperimentally test our approach on Amazon EC2, and show that the\ngeneralization error of approximate gradient coding is very close to the full\ngradient while requiring significantly less computation from the workers.\n",
"title": "Gradient Coding from Cyclic MDS Codes and Expander Graphs"
}
| null | null | null | null | true | null |
10949
| null |
Default
| null | null |
null |
{
"abstract": " The 10 MeV accelerator-driven subcritical system (ADS) Injector-I test stand\nat Institute of High Energy Physics (IHEP) is a testing facility dedicated to\ndemonstrate one of the two injector design schemes [Injector Scheme-I, which\nworks at 325 MHz], for the ADS project in China. The Injector adopted a four\nvane copper structure RFQ with output energy of 3.2 MeV and a superconducting\n(SC) section accommodating fourteen \\b{eta}g=0.12 single spoke cavities,\nfourteen SC solenoids and fourteen cold BPMs. The ion source was installed\nsince April of 2014, periods of commissioning are regularly scheduled between\ninstallation phases of the rest of the injector. Continuous wave (CW) beam was\nshooting through the injector and 10 MeV CW proton beam with average beam\ncurrent around 2 mA was obtained recently. This contribution describe the\nresults achieved so far and the difficulties encountered in CW commissioning.\n",
"title": "Commissioning of te China-ADS injector-I testing facility"
}
| null | null | null | null | true | null |
10950
| null |
Default
| null | null |
null |
{
"abstract": " Changes in the capital structure before and after the global financial crisis\nfor SMEs are studied, emphasizing their financing problems, distinguishing\nbetween internal financing and external financing determinants. The empirical\nresearch bears upon 158 small and medium-sized firms listed on Shenzhen and\nShanghai Stock Exchanges in China over the period of 2004-2014. A regression\nanalysis, along the lines of the Trade-Off Theory, shows that the leverage\ndecreases with profitability, non-debt tax shields and the liquidity, and\nincreases with firm size and tangibility. A positive relationship is found\nbetween firm growth and debt ratio, though not highly significantly. It is\nshown that the SMEs with high growth rates are those which will more easily\nobtain external financing after a financial crisis. It is recognized that the\nChina government should reconsider SMEs taxation laws.\n",
"title": "Impact of the Global Crisis on SME Internal vs. External Financing in China"
}
| null | null |
[
"Statistics"
] | null | true | null |
10951
| null |
Validated
| null | null |
null |
{
"abstract": " The evanescent field surrounding nano-scale optical waveguides offers an\nefficient interface between light and mesoscopic ensembles of neutral atoms.\nHowever, the thermal motion of trapped atoms, combined with the strong radial\ngradients of the guided light, leads to a time-modulated coupling between atoms\nand the light mode, thus giving rise to additional noise and motional dephasing\nof collective states. Here, we present a dipole force free scheme for coupling\nof the radial motional states, utilizing the strong intensity gradient of the\nguided mode and demonstrate all-optical coupling of the cesium hyperfine ground\nstates and motional sideband transitions. We utilize this to prolong the trap\nlifetime of an atomic ensemble by Raman sideband cooling of the radial motion,\nwhich has not been demonstrated in nano-optical structures previously. Our work\npoints towards full and independent control of internal and external atomic\ndegrees of freedom using guided light modes only.\n",
"title": "Dipole force free optical control and cooling of nanofiber trapped atoms"
}
| null | null | null | null | true | null |
10952
| null |
Default
| null | null |
null |
{
"abstract": " The analysis of industrial processes, modelled as descriptor systems, is\noften computationally hard due to the presence of both algebraic couplings and\ndifference equations of high order. In this paper, we introduce a control\nrefinement notion for these descriptor systems that enables analysis and\ncontrol design over related reduced-order systems. Utilising the behavioural\nframework, we extend upon the standard hierarchical control refinement for\nordinary systems and allow for algebraic couplings inherent to descriptor\nsystems.\n",
"title": "Control refinement for discrete-time descriptor systems: a behavioural approach via simulation relations"
}
| null | null | null | null | true | null |
10953
| null |
Default
| null | null |
null |
{
"abstract": " Increasing proton beam power on neutrino production targets is one of the\nmajor goals of the Fermilab long term accelerator programs. In this effort, the\nFermilab 8 GeV Booster synchrotron plays a critical role for at least the next\ntwo decades. Therefore, understanding the Booster in great detail is important\nas we continue to improve its performance. For example, it is important to know\naccurately the available RF power in the Booster by carrying out beam-based\nmeasurements in order to specify the needed upgrades to the Booster RF system.\nSince the Booster magnetic field is changing continuously measuring/calibrating\nthe RF voltage is not a trivial task. Here, we present a beam based method for\nthe RF voltage measurements. Data analysis is carried out using computer\nprograms developed in Python and MATLAB. The method presented here is\napplicable to any RCS which do not have flat-bottom and flat-top in the\nacceleration magnetic ramps. We have also carried out longitudinal beam\ntomography at injection and extraction energies with the data used for RF\nvoltage measurements. Beam based RF voltage measurements and beam tomography\nwere never done before for the Fermilab Booster. The results from these\ninvestigations will be very useful in future intensity upgrades.\n",
"title": "Beam Based RF Voltage Measurements and Longitudinal Beam Tomography at the Fermilab Booster"
}
| null | null | null | null | true | null |
10954
| null |
Default
| null | null |
null |
{
"abstract": " This paper presents our approach to the quantitative modeling and analysis of\nhighly (re)configurable systems, such as software product lines. Different\ncombinations of the optional features of such a system give rise to\ncombinatorially many individual system variants. We use a formal modeling\nlanguage that allows us to model systems with probabilistic behavior, possibly\nsubject to quantitative feature constraints, and able to dynamically install,\nremove or replace features. More precisely, our models are defined in the\nprobabilistic feature-oriented language QFLAN, a rich domain specific language\n(DSL) for systems with variability defined in terms of features. QFLAN\nspecifications are automatically encoded in terms of a process algebra whose\noperational behavior interacts with a store of constraints, and hence allows to\nseparate system configuration from system behavior. The resulting probabilistic\nconfigurations and behavior converge seamlessly in a semantics based on\ndiscrete-time Markov chains, thus enabling quantitative analysis. Our analysis\nis based on statistical model checking techniques, which allow us to scale to\nlarger models with respect to precise probabilistic analysis techniques. The\nanalyses we can conduct range from the likelihood of specific behavior to the\nexpected average cost, in terms of feature attributes, of specific system\nvariants. Our approach is supported by a novel Eclipse-based tool which\nincludes state-of-the-art DSL utilities for QFLAN based on the Xtext framework\nas well as analysis plug-ins to seamlessly run statistical model checking\nanalyses. We provide a number of case studies that have driven and validated\nthe development of our framework.\n",
"title": "A framework for quantitative modeling and analysis of highly (re)configurable systems"
}
| null | null | null | null | true | null |
10955
| null |
Default
| null | null |
null |
{
"abstract": " Many protostellar gapped and binary discs show misalignments between their\ninner and outer discs; in some cases, $\\sim70$ degree misalignments have been\nobserved. Here we show that these misalignments can be generated through a\n\"secular precession resonance\" between the nodal precession of the inner disc\nand the precession of the gap-opening (stellar or massive planetary) companion.\nAn evolving protostellar system may naturally cross this resonance during its\nlifetime due to disc dissipation and/or companion migration. If resonance\ncrossing occurs on the right timescale, of order a few Myrs, characteristic for\nyoung protostellar systems, the inner and outer discs can become highly\nmisaligned ($\\gtrsim 60$ degrees). When the primary star has a mass of order a\nsolar mass, generating a significant misalignment typically requires the\ncompanion to have a mass of $\\sim 0.01-0.1$ M$_\\odot$ and an orbital separation\nof tens of AU. The recently observed companion in the cavity of the gapped,\nhighly misaligned system HD 142527 satisfies these requirements, indicating\nthat a previous resonance crossing event misaligned the inner and outer discs.\nOur scenario for HD 142527's misaligned discs predicts that the companion's\norbital plane is aligned with the outer disc's; this prediction should be\ntestable with future observations as the companion's orbit is mapped out.\nMisalignments observed in several other gapped disc systems could be generated\nby the same secular resonance mechanism.\n",
"title": "Generating large misalignments in gapped and binary discs"
}
| null | null | null | null | true | null |
10956
| null |
Default
| null | null |
null |
{
"abstract": " We study a supersymmetric version of the Gardner equation (both focusing and\ndefocusing) using the superbilinear formalism. This equation is new and cannot\nbe obtained from supersymmetric modified Korteweg-de Vries equation with a\nnonzero boundary condition. We construct supersymmetric solitons and then by\npassing to the long-wave limit in the focusing case obtain rational nonsingular\nsolutions. We also discuss the supersymmetric version of the defocusing\nequation and the dynamics of its solutions.\n",
"title": "Bilinear approach to the supersymmetric Gardner equation"
}
| null | null | null | null | true | null |
10957
| null |
Default
| null | null |
null |
{
"abstract": " Many augmented reality (AR) applications operate within near-field reaching\ndistances, and require matching the depth of a virtual object with a real\nobject. The accuracy of this matching was measured in three experiments, which\nexamined the effect of focal distance, age, and brightness, within distances of\n33.3 to 50 cm, using a custom-built AR haploscope. Experiment I examined the\neffect of focal demand, at the levels of collimated (infinite focal distance),\nconsistent with other depth cues, and at the midpoint of reaching distance.\nObservers were too young to exhibit age-related reductions in accommodative\nability. The depth matches of collimated targets were increasingly\noverestimated with increasing distance, consistent targets were slightly\nunderestimated, and midpoint targets were accurately estimated. Experiment II\nreplicated Experiment I, with older observers. Results were similar to\nExperiment I. Experiment III replicated Experiment I with dimmer targets, using\nyoung observers. Results were again consistent with Experiment I, except that\nboth consistent and midpoint targets were accurately estimated. In all cases,\ncollimated results were explained by a model, where the collimation biases the\neyes' vergence angle outwards by a constant amount. Focal demand and brightness\naffect near-field AR depth matching, while age-related reductions in\naccommodative ability have no effect.\n",
"title": "The Effect of Focal Distance, Age, and Brightness on Near-Field Augmented Reality Depth Matching"
}
| null | null |
[
"Computer Science"
] | null | true | null |
10958
| null |
Validated
| null | null |
null |
{
"abstract": " The problem for two-dimensional steady water waves with vorticity is\nconsidered. Using methods of spatial dynamics, we reduce the problem to a\nfinite dimensional Hamiltonian system. As an application, we prove the\nexistence of non-symmetric steady water waves when the number of roots of the\ndispersion equation is greater than 1.\n",
"title": "Small-amplitude steady water waves with critical layers: non-symmetric waves"
}
| null | null |
[
"Mathematics"
] | null | true | null |
10959
| null |
Validated
| null | null |
null |
{
"abstract": " We study a neuro-inspired model that mimics a discussion (or information\ndissemination) process in a network of agents. During their interaction, agents\nredistribute activity and network weights, resulting in emergence of leader(s).\nThe model is able to reproduce the basic scenarios of leadership known in\nnature and society: laissez-faire (irregular activity, weak leadership, sizable\ninter-follower interaction, autonomous sub-leaders); participative or\ndemocratic (strong leadership, but with feedback from followers); and\nautocratic (no feedback, one-way influence). Several pertinent aspects of these\nscenarios are found as well---e.g., hidden leadership (a hidden clique of\nagents driving the official autocratic leader), and successive leadership (two\nleaders influence followers by turns). We study how these scenarios emerge from\ninter-agent dynamics and how they depend on behavior rules of agents---in\nparticular, on their inertia against state changes.\n",
"title": "Emergence of Leadership in Communication"
}
| null | null | null | null | true | null |
10960
| null |
Default
| null | null |
null |
{
"abstract": " In this article, we study the problem of controlling a highway segment facing\nstochastic perturbations, such as recurrent incidents and moving bottlenecks.\nTo model traffic flow under perturbations, we use the cell-transmission model\nwith Markovian capacities. The control inputs are: (i) the inflows that are\nsent to various on-ramps to the highway (for managing traffic demand), and (ii)\nthe priority levels assigned to the on-ramp traffic relative to the mainline\ntraffic (for allocating highway capacity). The objective is to maximize the\nthroughput while ensuring that on-ramp queues remain bounded in the long-run.\nWe develop a computational approach to solving this stability-constrained,\nthroughput-maximization problem. Firstly, we use the classical drift condition\nin stability analysis of Markov processes to derive a sufficient condition for\nboundedness of on-ramp queues. Secondly, we show that our control design\nproblem can be formulated as a mixed integer program with linear or bilinear\nconstraints, depending on the complexity of Lyapunov function involved in the\nstability condition. Finally, for specific types of capacity perturbations, we\nderive intuitive criteria for managing demand and/or selecting priority levels.\nThese criteria suggest that inflows and priority levels should be determined\nsimultaneously such that traffic queues are placed at locations that discharge\nqueues fast. We illustrate the performance benefits of these criteria through a\ncomputational study of a segment on Interstate 210 in California, USA.\n",
"title": "Throughput-Improving Control of Highways Facing Stochastic Perturbations"
}
| null | null | null | null | true | null |
10961
| null |
Default
| null | null |
null |
{
"abstract": " In portable, 3-D, or ultra-fast ultrasound (US) imaging systems, there is an\nincreasing demand to reconstruct high quality images from limited number of\ndata. However, the existing solutions require either hardware changes or\ncomputationally expansive algorithms. To overcome these limitations, here we\npropose a novel deep learning approach that interpolates the missing RF data by\nutilizing the sparsity of the RF data in the Fourier domain. Extensive\nexperimental results from sub-sampled RF data from a real US system confirmed\nthat the proposed method can effectively reduce the data rate without\nsacrificing the image quality.\n",
"title": "Deep Learning for Accelerated Ultrasound Imaging"
}
| null | null | null | null | true | null |
10962
| null |
Default
| null | null |
null |
{
"abstract": " This paper extends a conventional, general framework for online adaptive\nestimation problems for systems governed by unknown nonlinear ordinary\ndifferential equations. The central feature of the theory introduced in this\npaper represents the unknown function as a member of a reproducing kernel\nHilbert space (RKHS) and defines a distributed parameter system (DPS) that\ngoverns state estimates and estimates of the unknown function. This paper 1)\nderives sufficient conditions for the existence and stability of the infinite\ndimensional online estimation problem, 2) derives existence and stability of\nfinite dimensional approximations of the infinite dimensional approximations,\nand 3) determines sufficient conditions for the convergence of finite\ndimensional approximations to the infinite dimensional online estimates. A new\ncondition for persistency of excitation in a RKHS in terms of its evaluation\nfunctionals is introduced in the paper that enables proof of convergence of the\nfinite dimensional approximations of the unknown function in the RKHS. This\npaper studies two particular choices of the RKHS, those that are generated by\nexponential functions and those that are generated by multiscale kernels\ndefined from a multiresolution analysis.\n",
"title": "Adaptive Estimation for Nonlinear Systems using Reproducing Kernel Hilbert Spaces"
}
| null | null | null | null | true | null |
10963
| null |
Default
| null | null |
null |
{
"abstract": " The two model-theoretic concepts of weak saturation and weak amalgamation\nproperty are studied in the context of accessible categories. We relate these\ntwo concepts providing sufficient conditions for existence and uniqueness of\nweakly saturated objects of an accessible category K. We discuss the\nimplications of this fact in classical model theory.\n",
"title": "Weak saturation and weak amalgamation property"
}
| null | null | null | null | true | null |
10964
| null |
Default
| null | null |
null |
{
"abstract": " This paper is concerned with the behavior of the ergodic constant associated\nwith convex and superlinear Hamilton-Jacobi equation in a periodic environment\nwhich is perturbed either by medium with increasing period or by a random\nBernoulli perturbation with small parameter. We find a first order Taylor's\nexpansion for the ergodic constant which depends on the dimension d. When d = 1\nthe first order term is non trivial, while for all d $\\ge$ 2 it is always 0.\nAlthough such questions have been looked at in the context of linear uniformly\nelliptic homogenization, our results are the first of this kind in nonlinear\nsettings. Our arguments, which rely on viscosity solutions and the weak KAM\ntheory, also raise several new and challenging questions.\n",
"title": "Perturbation problems in homogenization of hamilton-jacobi equations"
}
| null | null | null | null | true | null |
10965
| null |
Default
| null | null |
null |
{
"abstract": " Collective behavior among coupled dynamical units can emerge in various forms\nas a result of different coupling topologies as well as different types of\ncoupling functions. Chimera states have recently received ample attention as a\nfascinating manifestation of collective behavior, in particular describing a\nsymmetry breaking spatiotemporal pattern where synchronized and desynchronized\nstates coexist in a network of coupled oscillators. In this perspective, we\nreview the emergence of different chimera states, focusing on the effects of\ndifferent coupling topologies that describe the interaction network connecting\nthe oscillators. We cover chimera states that emerge in local, nonlocal and\nglobal coupling topologies, as well as in modular, temporal and multilayer\nnetworks. We also provide an outline of challenges and directions for future\nresearch.\n",
"title": "Chimera states: Effects of different coupling topologies"
}
| null | null | null | null | true | null |
10966
| null |
Default
| null | null |
null |
{
"abstract": " Human activity recognition using smart home sensors is one of the bases of\nubiquitous computing in smart environments and a topic undergoing intense\nresearch in the field of ambient assisted living. The increasingly large amount\nof data sets calls for machine learning methods. In this paper, we introduce a\ndeep learning model that learns to classify human activities without using any\nprior knowledge. For this purpose, a Long Short Term Memory (LSTM) Recurrent\nNeural Network was applied to three real world smart home datasets. The results\nof these experiments show that the proposed approach outperforms the existing\nones in terms of accuracy and performance.\n",
"title": "Human Activity Recognition using Recurrent Neural Networks"
}
| null | null | null | null | true | null |
10967
| null |
Default
| null | null |
null |
{
"abstract": " We consider eigenvalue problems for elliptic operators of arbitrary order\n$2m$ subject to Neumann boundary conditions on bounded domains of the Euclidean\n$N$-dimensional space. We study the dependence of the eigenvalues upon\nvariations of mass density and in particular we discuss the existence and\ncharacterization of upper and lower bounds under both the condition that the\ntotal mass is fixed and the condition that the $L^{\\frac{N}{2m}}$-norm of the\ndensity is fixed. We highlight that the interplay between the order of the\noperator and the space dimension plays a crucial role in the existence of\neigenvalue bounds.\n",
"title": "Eigenvalues of elliptic operators with density"
}
| null | null | null | null | true | null |
10968
| null |
Default
| null | null |
null |
{
"abstract": " This article develops a framework for testing general hypothesis in\nhigh-dimensional models where the number of variables may far exceed the number\nof observations. Existing literature has considered less than a handful of\nhypotheses, such as testing individual coordinates of the model parameter.\nHowever, the problem of testing general and complex hypotheses remains widely\nopen. We propose a new inference method developed around the hypothesis\nadaptive projection pursuit framework, which solves the testing problems in the\nmost general case. The proposed inference is centered around a new class of\nestimators defined as $l_1$ projection of the initial guess of the unknown onto\nthe space defined by the null. This projection automatically takes into account\nthe structure of the null hypothesis and allows us to study formal inference\nfor a number of long-standing problems. For example, we can directly conduct\ninference on the sparsity level of the model parameters and the minimum signal\nstrength. This is especially significant given the fact that the former is a\nfundamental condition underlying most of the theoretical development in\nhigh-dimensional statistics, while the latter is a key condition used to\nestablish variable selection properties. Moreover, the proposed method is\nasymptotically exact and has satisfactory power properties for testing very\ngeneral functionals of the high-dimensional parameters. The simulation studies\nlend further support to our theoretical claims and additionally show excellent\nfinite-sample size and power properties of the proposed test.\n",
"title": "A projection pursuit framework for testing general high-dimensional hypothesis"
}
| null | null | null | null | true | null |
10969
| null |
Default
| null | null |
null |
{
"abstract": " Quantum computing technologies have become a hot topic in academia and\nindustry receiving much attention and financial support from all sides.\nBuilding a quantum computer that can be used practically is in itself an\noutstanding challenge that has become the 'new race to the moon'. Next to\nresearchers and vendors of future computing technologies, national authorities\nare showing strong interest in maturing this technology due to its known\npotential to break many of today's encryption techniques, which would have\nsignificant impact on our society. It is however quite likely that quantum\ncomputing has beneficial impact on many computational disciplines.\nIn this article we describe our vision of future developments in scientific\ncomputing that would be enabled by the advent of software-programmable quantum\ncomputers. We thereby assume that quantum computers will form part of a hybrid\naccelerated computing platform like GPUs and co-processor cards do today. In\nparticular, we address the potential of quantum algorithms to bring major\nbreakthroughs in applied mathematics and its applications. Finally, we give\nseveral examples that demonstrate the possible impact of quantum-accelerated\nscientific computing on society.\n",
"title": "On the impact of quantum computing technology on future developments in high-performance scientific computing"
}
| null | null | null | null | true | null |
10970
| null |
Default
| null | null |
null |
{
"abstract": " Recently, encoder-decoder neural networks have shown impressive performance\non many sequence-related tasks. The architecture commonly uses an attentional\nmechanism which allows the model to learn alignments between the source and the\ntarget sequence. Most attentional mechanisms used today is based on a global\nattention property which requires a computation of a weighted summarization of\nthe whole input sequence generated by encoder states. However, it is\ncomputationally expensive and often produces misalignment on the longer input\nsequence. Furthermore, it does not fit with monotonous or left-to-right nature\nin several tasks, such as automatic speech recognition (ASR),\ngrapheme-to-phoneme (G2P), etc. In this paper, we propose a novel attention\nmechanism that has local and monotonic properties. Various ways to control\nthose properties are also explored. Experimental results on ASR, G2P and\nmachine translation between two languages with similar sentence structures,\ndemonstrate that the proposed encoder-decoder model with local monotonic\nattention could achieve significant performance improvements and reduce the\ncomputational complexity in comparison with the one that used the standard\nglobal attention architecture.\n",
"title": "Local Monotonic Attention Mechanism for End-to-End Speech and Language Processing"
}
| null | null | null | null | true | null |
10971
| null |
Default
| null | null |
null |
{
"abstract": " Approximate dynamic programming algorithms, such as approximate value\niteration, have been successfully applied to many complex reinforcement\nlearning tasks, and a better approximate dynamic programming algorithm is\nexpected to further extend the applicability of reinforcement learning to\nvarious tasks. In this paper we propose a new, robust dynamic programming\nalgorithm that unifies value iteration, advantage learning, and dynamic policy\nprogramming. We call it generalized value iteration (GVI) and its approximated\nversion, approximate GVI (AGVI). We show AGVI's performance guarantee, which\nincludes performance guarantees for existing algorithms, as special cases. We\ndiscuss theoretical weaknesses of existing algorithms, and explain the\nadvantages of AGVI. Numerical experiments in a simple environment support\ntheoretical arguments, and suggest that AGVI is a promising alternative to\nprevious algorithms.\n",
"title": "Unifying Value Iteration, Advantage Learning, and Dynamic Policy Programming"
}
| null | null | null | null | true | null |
10972
| null |
Default
| null | null |
null |
{
"abstract": " Spatiotemporal forecasting has various applications in neuroscience, climate\nand transportation domain. Traffic forecasting is one canonical example of such\nlearning task. The task is challenging due to (1) complex spatial dependency on\nroad networks, (2) non-linear temporal dynamics with changing road conditions\nand (3) inherent difficulty of long-term forecasting. To address these\nchallenges, we propose to model the traffic flow as a diffusion process on a\ndirected graph and introduce Diffusion Convolutional Recurrent Neural Network\n(DCRNN), a deep learning framework for traffic forecasting that incorporates\nboth spatial and temporal dependency in the traffic flow. Specifically, DCRNN\ncaptures the spatial dependency using bidirectional random walks on the graph,\nand the temporal dependency using the encoder-decoder architecture with\nscheduled sampling. We evaluate the framework on two real-world large scale\nroad network traffic datasets and observe consistent improvement of 12% - 15%\nover state-of-the-art baselines.\n",
"title": "Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting"
}
| null | null | null | null | true | null |
10973
| null |
Default
| null | null |
null |
{
"abstract": " Recently, machine learning has been used in every possible field to leverage\nits amazing power. For a long time, the net-working and distributed computing\nsystem is the key infrastructure to provide efficient computational resource\nfor machine learning. Networking itself can also benefit from this promising\ntechnology. This article focuses on the application of Machine Learning\ntechniques for Networking (MLN), which can not only help solve the intractable\nold network questions but also stimulate new network applications. In this\narticle, we summarize the basic workflow to explain how to apply the machine\nlearning technology in the networking domain. Then we provide a selective\nsurvey of the latest representative advances with explanations on their design\nprinciples and benefits. These advances are divided into several network design\nobjectives and the detailed information of how they perform in each step of MLN\nworkflow is presented. Finally, we shed light on the new opportunities on\nnetworking design and community building of this new inter-discipline. Our goal\nis to provide a broad research guideline on networking with machine learning to\nhelp and motivate researchers to develop innovative algorithms, standards and\nframeworks.\n",
"title": "Machine Learning for Networking: Workflow, Advances and Opportunities"
}
| null | null | null | null | true | null |
10974
| null |
Default
| null | null |
null |
{
"abstract": " We present the WiFeS Atlas of Galactic Globular cluster Spectra, a library of\nintegrated spectra of Milky Way and Local Group globular clusters. We used the\nWiFeS integral field spectrograph on the Australian National University 2.3 m\ntelescope to observe the central regions of 64 Milky Way globular clusters and\n22 globular clusters hosted by the Milky Way's low mass satellite galaxies. The\nspectra have wider wavelength coverage (3300 {\\AA} to 9050 {\\AA}) and higher\nspectral resolution (R = 6800) than existing spectral libraries of Milky Way\nglobular clusters. By including Large and Small Magellanic Cloud star clusters,\nwe extend the coverage of parameter space of existing libraries towards young\nand intermediate ages. While testing stellar population synthesis models and\nanalysis techniques is the main aim of this library, the observations may also\nfurther our understanding of the stellar populations of Local Group globular\nclusters and make possible the direct comparison of extragalactic globular\ncluster integrated light observations with well understood globular clusters in\nthe Milky Way. The integrated spectra are publicly available via the project\nwebsite.\n",
"title": "The WAGGS project - I. The WiFeS Atlas of Galactic Globular cluster Spectra"
}
| null | null | null | null | true | null |
10975
| null |
Default
| null | null |
null |
{
"abstract": " The spread of new products in a networked population is often modeled as an\nepidemic. However, in the case of \"complex\" contagion, these models are\ninsufficient to properly model adoption behavior. In this paper, we investigate\na model of complex contagion which allows a coevolutionary interplay between\nadoption, modeled as an SIS epidemic spreading process, and social\nreinforcement effects, modeled as consensus opinion dynamics. Asymptotic\nstability analysis of the all-adopt as well as the none-adopt equilibria of the\ncombined opinion-adoption model is provided through the use of Lyapunov\narguments. In doing so, sufficient conditions are provided which determine the\nstability of the \"flop\" state, where no one adopts the product and everyone's\nopinion of the product is least favorable, and the \"hit\" state, where everyone\nadopts and their opinions are most favorable. These conditions are shown to\nextend to the bounded confidence opinion dynamic under a stronger assumption on\nthe model parameters. To conclude, numerical simulations demonstrate behavior\nof the model which reflect findings from the sociology literature on adoption\nbehavior.\n",
"title": "Going Viral: Stability of Consensus-Driven Adoptive Spread"
}
| null | null | null | null | true | null |
10976
| null |
Default
| null | null |
null |
{
"abstract": " We develop a linear algebraic framework for the shape-from-shading problem,\nbecause tensors arise when scalar (e.g. image) and vector (e.g. surface normal)\nfields are differentiated multiple times. The work is in two parts. In this\nfirst part we investigate when image derivatives exhibit invariance to changing\nillumination by calculating the statistics of image derivatives under general\ndistributions on the light source. We computationally validate the hypothesis\nthat image orientations (derivatives) provide increased invariance to\nillumination by showing (for a Lambertian model) that a shape-from-shading\nalgorithm matching gradients instead of intensities provides more accurate\nreconstructions when illumination is incorrectly estimated under a flatness\nprior.\n",
"title": "What's In A Patch, I: Tensors, Differential Geometry and Statistical Shading Analysis"
}
| null | null | null | null | true | null |
10977
| null |
Default
| null | null |
null |
{
"abstract": " We present a projectively invariant description of planar linear 3-webs and\nconstruct a counterexample to Gronwall's conjecture.\n",
"title": "Counterexample to Gronwall's Conjecture"
}
| null | null |
[
"Mathematics"
] | null | true | null |
10978
| null |
Validated
| null | null |
null |
{
"abstract": " Modern cities are growing ecosystems that face new challenges due to the\nincreasing population demands. One of the many problems they face nowadays is\nwaste management, which has become a pressing issue requiring new solutions.\nSwarm robotics systems have been attracting an increasing amount of attention\nin the past years and they are expected to become one of the main driving\nfactors for innovation in the field of robotics. The research presented in this\npaper explores the feasibility of a swarm robotics system in an urban\nenvironment. By using bio-inspired foraging methods such as multi-place\nforaging and stigmergy-based navigation, a swarm of robots is able to improve\nthe efficiency and autonomy of the urban waste management system in a realistic\nscenario. To achieve this, a diverse set of simulation experiments was\nconducted using real-world GIS data and implementing different garbage\ncollection scenarios driven by robot swarms. Results presented in this research\nshow that the proposed system outperforms current approaches. Moreover, results\nnot only show the efficiency of our solution, but also give insights about how\nto design and customize these systems.\n",
"title": "Urban Swarms: A new approach for autonomous waste management"
}
| null | null |
[
"Computer Science"
] | null | true | null |
10979
| null |
Validated
| null | null |
null |
{
"abstract": " There are so many vehicles in the world and the number of vehicles is\nincreasing rapidly. To alleviate the parking problems caused by that, the smart\nparking system has been developed. The parking planning is one of the most\nimportant parts of it. An effective parking planning strategy makes the better\nuse of parking resources possible. In this paper, we present a feasible method\nto do parking planning. We transform the parking planning problem into a kind\nof linear assignment problem. We take vehicles as jobs and parking spaces as\nagents. We take distances between vehicles and parking spaces as costs for\nagents doing jobs. Then we design an algorithm for this particular assignment\nproblem and solve the parking planning problem. The method proposed can give\ntimely and efficient guide information to vehicles for a real time smart\nparking system. Finally, we show the effectiveness of the method with\nexperiments over some data, which can simulate the situation of doing parking\nplanning in the real world.\n",
"title": "An Algorithm of Parking Planning for Smart Parking System"
}
| null | null | null | null | true | null |
10980
| null |
Default
| null | null |
null |
{
"abstract": " We describe categorical models of a circuit-based (quantum) functional pro-\ngramming language. We show that enriched categories play a crucial role.\nFollowing earlier work on QWire by Paykin et al., we consider both a simple\nfirst-order linear language for circuits, and a more powerful host language,\nsuch that the circuit language is embedded inside the host language. Our\ncategorical semantics for the host language is standard, and involves cartesian\nclosed categories and monads. We interpret the circuit language not in an\nordinary category, but in a category that is enriched in the host category. We\nshow that this structure is also related to linear/non-linear models. As an\nextended example, we recall an earlier result that the category of W*-algebras\nis dcpo-enriched, and we use this model to extend the circuit language with\nsome recursive types.\n",
"title": "Classical Control, Quantum Circuits and Linear Logic in Enriched Category Theory"
}
| null | null |
[
"Computer Science",
"Mathematics"
] | null | true | null |
10981
| null |
Validated
| null | null |
null |
{
"abstract": " This works presents a formulation for visual navigation that unifies map\nbased spatial reasoning and path planning, with landmark based robust plan\nexecution in noisy environments. Our proposed formulation is learned from data\nand is thus able to leverage statistical regularities of the world. This allows\nit to efficiently navigate in novel environments given only a sparse set of\nregistered images as input for building representations for space. Our\nformulation is based on three key ideas: a learned path planner that outputs\npath plans to reach the goal, a feature synthesis engine that predicts features\nfor locations along the planned path, and a learned goal-driven closed loop\ncontroller that can follow plans given these synthesized features. We test our\napproach for goal-driven navigation in simulated real world environments and\nreport performance gains over competitive baseline approaches.\n",
"title": "Unifying Map and Landmark Based Representations for Visual Navigation"
}
| null | null | null | null | true | null |
10982
| null |
Default
| null | null |
null |
{
"abstract": " Since the 1940s, population projections have in most cases been produced\nusing the deterministic cohort component method. However, in 2015, for the\nfirst time, in a major advance, the United Nations issued official\nprobabilistic population projections for all countries based on Bayesian\nhierarchical models for total fertility and life expectancy. The estimates of\nthese models and the resulting projections are conditional on the UN's official\nestimates of past values. However, these past values are themselves uncertain,\nparticularly for the majority of the world's countries that do not have\nlongstanding high-quality vital registration systems, when they rely on surveys\nand censuses with their own biases and measurement errors. This paper is a\nfirst attempt to remedy this for total fertility rates, by extending the UN\nmodel for the future to take account of uncertainty about past values. This is\ndone by adding an additional level to the hierarchical model to represent the\nmultiple data sources, in each case estimating their bias and measurement error\nvariance. We assess the method by out-of-sample predictive validation. While\nthe prediction intervals produced by the current method have somewhat less than\nnominal coverage, we find that our proposed method achieves close to nominal\ncoverage. The prediction intervals become wider for countries for which the\nestimates of past total fertility rates rely heavily on surveys rather than on\nvital registration data.\n",
"title": "Accounting for Uncertainty About Past Values In Probabilistic Projections of the Total Fertility Rate for All Countries"
}
| null | null | null | null | true | null |
10983
| null |
Default
| null | null |
null |
{
"abstract": " The existence of elliptic periodic solutions of a perturbed Kepler problem is\nproved. The equations are in the plane and the perturbation depends\nperiodically on time. The proof is based on a local description of the\nsymplectic group in two degrees of freedom.\n",
"title": "Periodic solutions of a perturbed Kepler problem in the plane: from existence to stability"
}
| null | null | null | null | true | null |
10984
| null |
Default
| null | null |
null |
{
"abstract": " Suppression of interference from narrowband frequency signals play vital role\nin many signal processing and communication applications. A transform based\nmethod for suppression of narrow band interference in a biomedical signal is\nproposed. As a specific example Electrocardiogram (ECG) is considered for the\nanalysis. ECG is one of the widely used biomedical signal. ECG signal is often\ncontaminated with baseline wander noise, powerline interference (PLI) and\nartifacts (bioelectric signals), which complicates the processing of raw ECG\nsignal. This work proposes an approach using Ramanujan periodic transform for\nreducing PLI and is tested on a subject data from MIT-BIH Arrhythmia database.\nA sum ($E$) of Euclidean error per block ($e_i$) is used as measure to quantify\nthe suppression capability of RPT and notch filter based methods. The\ntransformation is performed for different lengths ($N$), namely $36$, $72$,\n$108$, $144$, $180$. Every doubling of $N$-points results in $50{\\%}$ reduction\nin error ($E$).\n",
"title": "Removal of Narrowband Interference (PLI in ECG Signal) Using Ramanujan Periodic Transform (RPT)"
}
| null | null | null | null | true | null |
10985
| null |
Default
| null | null |
null |
{
"abstract": " It is well known that the Euler vortex patch in $\\mathbb{R}^{2}$ will remain\nregular if it is regular enough initially. In bounded domains, the regularity\ntheory for patch solutions is less complete. We study here the Euler vortex\npatch in a disk. We prove global in time regularity by providing the upper\nbound of the growth of curvature of the patch boundary. For a special symmetric\nscenario, we construct an example of double exponential curvature growth,\nshowing that such upper bound is qualitatively sharp.\n",
"title": "Global regularity and fast small scale formation for Euler patch equation in a disk"
}
| null | null | null | null | true | null |
10986
| null |
Default
| null | null |
null |
{
"abstract": " The effective interaction between the itinerant spin degrees of freedom in\nthe paramagnetic phases of hole doped quantum Heisenberg antiferromagnets is\ninvestigated theoretically, based on the single-band t-J model on 1D lattice,\nat zero temperature. The effective spin-spin interaction for this model in the\nstrong correlation limit, is studied in terms of the generalized spin stiffness\nconstant as a function of doping concentration. The plot of this generalized\nspin stiffness constant against doping shows a very high value of stiffness in\nthe vicinity of zero doping and a very sharp fall with increase in doping\nconcentration, signifying the rapid decay of original coupling of\nsemi-localized spins in the system. Quite interestingly, this plot also shows a\nmaximum occurring at a finite value of doping, which strongly suggests the\ntendency of the itinerant spins to couple again in the unconventional\nparamagnetic phase. As the doping is further increased, this new coupling is\nalso suppressed and the spin response becomes analogous to almost Pauli-like.\nThe last two predictions of ours are quite novel and may be directly tested by\nindependent experiments and computational techniques in future. Our results in\ngeneral receive good support from other theoretical works and experimental\nresults extracted from the chains of YBa$_2$Cu$_3$O$_{6+x}$.\n",
"title": "Effective interaction in a non-Fermi liquid conductor and spin correlations in under-doped cuprates"
}
| null | null | null | null | true | null |
10987
| null |
Default
| null | null |
null |
{
"abstract": " Human collaborators coordinate effectively their actions through both verbal\nand non-verbal communication. We believe that the the same should hold for\nhuman-robot teams. We propose a formalism that enables a robot to decide\noptimally between doing a task and issuing an utterance. We focus on two types\nof utterances: verbal commands, where the robot expresses how it wants its\nhuman teammate to behave, and state-conveying actions, where the robot explains\nwhy it is behaving this way. Human subject experiments show that enabling the\nrobot to issue verbal commands is the most effective form of communicating\nobjectives, while retaining user trust in the robot. Communicating why\ninformation should be done judiciously, since many participants questioned the\ntruthfulness of the robot statements.\n",
"title": "Planning with Verbal Communication for Human-Robot Collaboration"
}
| null | null | null | null | true | null |
10988
| null |
Default
| null | null |
null |
{
"abstract": " We consider the phenomenon of Bose-Einstein condensation of quasi-equilibrium\nmagnons which leads to a spin superfluidity, the coherent quantum transfer of\nmagnetization in magnetic materials. These phenomena are beyond the classical\nLandau-Lifshitz-Gilbert paradigm. The critical conditions for excited magnon\ndensity for ferro- and antiferromagnets, bulk and thin films are estimated and\ndiscussed. The BEC should occur in the antiferromagnetic hematite at much lower\nexcited magnon density compared to the ferromagnetic YIG.\n",
"title": "Magnon Condensation and Spin Superfluidity"
}
| null | null |
[
"Physics"
] | null | true | null |
10989
| null |
Validated
| null | null |
null |
{
"abstract": " We study the stability of the electroweak vacuum in low-scale inflation\nmodels whose Hubble parameter is much smaller than the instability scale of the\nHiggs potential. In general, couplings between the inflaton and Higgs are\npresent, and hence we study effects of these couplings during and after\ninflation. We derive constraints on the couplings between the inflaton and\nHiggs by requiring that they do not lead to catastrophic electroweak vacuum\ndecay, in particular, via resonant production of the Higgs particles.\n",
"title": "Electroweak Vacuum Metastability and Low-scale Inflation"
}
| null | null | null | null | true | null |
10990
| null |
Default
| null | null |
null |
{
"abstract": " We explore the spectral properties of a capillary dye laser in the highly\nmultimode regime. Our experiments indicate that the spectral behavior of the\nlaser does not conform with a simple Fabry-Perot analysis; rather, it is\nstrongly dictated by a Vernier resonant mechanism involving multiple modes,\nwhich propagate with different group velocities. The laser operates over a very\nbroad spectral range and the Vernier effect gives rise to a free spectral range\nwhich is orders of magnitude larger than that expected from a simple\nFabry-Perot mechanism. The presented theoretical calculations confirm the\nexperimental results. Propagating modes of the capillary fiber are calculated\nusing the finite element method (FEM) and it is shown that the optical\npathlengths resulting from simultaneous beatings of these modes are in close\nagreement with the optical pathlengths directly extracted from the Fourier\nTransform of the experimentally measured laser emission spectra.\n",
"title": "Spectral selectivity in capillary dye lasers"
}
| null | null | null | null | true | null |
10991
| null |
Default
| null | null |
null |
{
"abstract": " Electrolyte gating is widely used to induce large carrier density modulation\non solid surfaces to explore various properties. Most of past works have\nattributed the charge modulation to electrostatic field effect. However, some\nrecent reports have argued that the electrolyte gating effect in VO2, TiO2 and\nSrTiO3 originated from field-induced oxygen vacancy formation. This gives rise\nto a controversy about the gating mechanism, and it is therefore vital to\nreveal the relationship between the role of electrolyte gating and the\nintrinsic properties of materials. Here, we report entirely different\nmechanisms of electrolyte gating on two high-Tc cuprates, NdBa2Cu3O7-{\\delta}\n(NBCO) and Pr2-xCexCuO4 (PCCO), with different crystal structures. We show that\nfield-induced oxygen vacancy formation in CuO chains of NBCO plays the dominant\nrole while it is mainly an electrostatic field effect in the case of PCCO. The\npossible reason is that NBCO has mobile oxygen in CuO chains while PCCO does\nnot. Our study helps clarify the controversy relating to the mechanism of\nelectrolyte gating, leading to a better understanding of the role of oxygen\nelectro migration which is very material specific.\n",
"title": "The Mechanism of Electrolyte Gating on High-Tc Cuprates: The Role of Oxygen Migration and Electrostatics"
}
| null | null |
[
"Physics"
] | null | true | null |
10992
| null |
Validated
| null | null |
null |
{
"abstract": " In the past few years, an action of $\\mathrm{PGL}_2(\\mathbb F_q)$ on the set\nof irreducible polynomials in $\\mathbb F_q[x]$ has been introduced and many\nquestions have been discussed, such as the characterization and number of\ninvariant elements. In this paper, we analyze some recent works on this action\nand provide full generalizations of them, yielding final theoretical results on\nthe characterization and number of invariant elements.\n",
"title": "Invariant theory of a special group action on irreducible polynomials over finite fields"
}
| null | null |
[
"Mathematics"
] | null | true | null |
10993
| null |
Validated
| null | null |
null |
{
"abstract": " We provide a novel notion of what it means to be interpretable, looking past\nthe usual association with human understanding. Our key insight is that\ninterpretability is not an absolute concept and so we define it relative to a\ntarget model, which may or may not be a human. We define a framework that\nallows for comparing interpretable procedures by linking them to important\npractical aspects such as accuracy and robustness. We characterize many of the\ncurrent state-of-the-art interpretable methods in our framework portraying its\ngeneral applicability. Finally, principled interpretable strategies are\nproposed and empirically evaluated on synthetic data, as well as on the largest\npublic olfaction dataset that was made recently available \\cite{olfs}. We also\nexperiment on MNIST with a simple target model and different oracle models of\nvarying complexity. This leads to the insight that the improvement in the\ntarget model is not only a function of the oracle model's performance, but also\nits relative complexity with respect to the target model. Further experiments\non CIFAR-10, a real manufacturing dataset and FICO dataset showcase the benefit\nof our methods over Knowledge Distillation when the target models are simple\nand the complex model is a neural network.\n",
"title": "TIP: Typifying the Interpretability of Procedures"
}
| null | null | null | null | true | null |
10994
| null |
Default
| null | null |
null |
{
"abstract": " In some laboratory and most astrophysical situations plasma wake-field\nacceleration of electrons is one dimensional, i.e. variation transverse to the\nbeam's motion can be ignored. Thus, one dimensional (1D), particle-in-cell\n(PIC), fully electromagnetic simulations of electron plasma wake field\nacceleration are conducted in order to study the differences in electron plasma\nwake field acceleration in MeV versus GeV and linear versus blowout regimes.\nFirst, we show that caution needs to be taken when using fluid simulations, as\nPIC simulations prove that an approximation for an electron bunch not to evolve\nin time for few hundred plasma periods only applies when it is sufficiently\nrelativistic. This conclusion is true irrespective of the plasma temperature.\nWe find that in the linear regime and GeV energies, the accelerating electric\nfield generated by the plasma wake is similar to the linear and MeV regime.\nHowever, because GeV energy driving bunch stays intact for much longer time,\nthe final acceleration energies are much larger in the GeV energies case. In\nthe GeV energy range and blowout regime the wake's accelerating electric field\nis much larger in amplitude compared to the linear case and also plasma wake\ngeometrical size is much larger. Thus, the correct positioning of the trailing\nbunch is needed to achieve the efficient acceleration. For the considered case,\noptimally there should be approximately $(90-100) c/\\omega_{pe}$ distance\nbetween trailing and driving electron bunches in the GeV blowout regime.\n",
"title": "Differences in 1D electron plasma wake field acceleration in MeV versus GeV and linear versus blowout regimes"
}
| null | null | null | null | true | null |
10995
| null |
Default
| null | null |
null |
{
"abstract": " Consider an undirected mixed membership network with $n$ nodes and $K$\ncommunities. For each node $1 \\leq i \\leq n$, we model the membership by\n$\\pi_{i} = (\\pi_{i}(1), \\pi_{i}(2), \\ldots$, $\\pi_{i}(K))'$, where $\\pi_{i}(k)$\nis the probability that node $i$ belongs to community $k$, $1 \\leq k \\leq K$.\nWe call node $i$ \"pure\" if $\\pi_i$ is degenerate and \"mixed\" otherwise. The\nprimary interest is to estimate $\\pi_i$, $1 \\leq i \\leq n$.\nWe model the adjacency matrix $A$ with a Degree Corrected Mixed Membership\n(DCMM) model. Let $\\hat{\\xi}_1, \\hat{\\xi}_2, \\ldots, \\hat{\\xi}_K$ be the first\n$K$ eigenvectors of $A$. We define a matrix $\\hat{R} \\in \\mathbb{R}^{n, K-1}$\nby $\\hat{R}(i,k) = \\hat{\\xi}_{k+1}(i)/\\hat{\\xi}_1(i)$, $1 \\leq k \\leq K-1$, $1\n\\leq i \\leq n$. The matrix can be viewed as a distorted version of its\nnon-stochastic counterpart $R \\in \\mathbb{R}^{n, K-1}$, which is unknown but\ncontains all information we need for the memberships.\nWe reveal an interesting insight: There is a simplex ${\\cal S}$ in\n$\\mathbb{R}^{K-1}$ such that row $i$ of $R$ corresponds to a vertex of ${\\cal\nS}$ if node $i$ is pure, and corresponds to an interior point of ${\\cal S}$\notherwise. Vertex Hunting (i.e., estimating the vertices of ${\\cal S}$) is thus\nthe key to our problem.\nThe matrix $\\hat{R}$ is a row-wise normalization on the matrix of\neigenvectors $\\hat{\\Xi}=[\\hat{\\xi}_1,\\ldots,\\hat{\\xi}_K]$, first proposed by\nJin (2015). Alternatively, we may normalize $\\hat{\\Xi}$ by the row-wise\n$\\ell^q$-norms (e.g., Supplement of Jin (2015)), but it won't give rise to a\nsimplex so is less convenient.\nWe propose a new approach $\\textit{Mixed-SCORE}$ to estimating the\nmemberships, at the heart of which is an easy-to-use Vertex Hunting algorithm.\nThe approach is successfully applied to $4$ network data sets. We also derive\nthe rate of convergence for Mixed-SCORE.\n",
"title": "Estimating network memberships by simplex vertex hunting"
}
| null | null | null | null | true | null |
10996
| null |
Default
| null | null |
null |
{
"abstract": " In this paper, we derive the non-singular Green's functions for the unbounded\nPoisson equation in two and three dimensions using a spectral approach to\nregularize the homogeneous equation. The resulting Green's functions are\nrelevant to applications which are restricted to a minimum resolved length\nscale (e.g. a mesh size h) and thus cannot handle the singular Green's function\nof the continuous Poisson equation. We furthermore derive the gradient vector\nof the regularized Green's function, as this is useful in applications where\nthe Poisson equation represents potential functions of a vector field.\n",
"title": "Non-singular Green's functions for the unbounded Poisson equation in one, two and three dimensions"
}
| null | null | null | null | true | null |
10997
| null |
Default
| null | null |
null |
{
"abstract": " We report a survey of molecular gas in galaxies in the XMMXCS J2215.9-1738\ncluster at $z=1.46$. We have detected emission lines from 17 galaxies within a\nradius of $R_{200}$ from the cluster center, in Band 3 data of the Atacama\nLarge Millimeter/submillimeter Array (ALMA) with a coverage of 93 -- 95 GHz in\nfrequency and 2.33 arcmin$^2$ in spatial direction. The lines are all\nidentified as CO $J$=2--1 emission lines from cluster members at $z\\sim1.46$ by\ntheir redshifts and the colors of their optical and near-infrared (NIR)\ncounterparts. The line luminosities reach down to $L'_{\\rm\nCO(2-1)}=4.5\\times10^{9}$ K km s$^{-1}$ pc$^2$. The spatial distribution of\ngalaxies with a detection of CO(2--1) suggests that they disappear from the\nvery center of the cluster. The phase-space diagram showing relative velocity\nversus cluster-centric distance indicates that the gas-rich galaxies have\nentered the cluster more recently than the gas-poor star-forming galaxies and\npassive galaxies located in the virialized region of this cluster. The results\nimply that the galaxies have experienced ram-pressure stripping and/or\nstrangulation during the course of infall towards the cluster center and then\nthe molecular gas in the galaxies at the cluster center is depleted by star\nformation.\n",
"title": "Evolutionary phases of gas-rich galaxies in a galaxy cluster at z=1.46"
}
| null | null | null | null | true | null |
10998
| null |
Default
| null | null |
null |
{
"abstract": " The iterative ensemble Kalman filter (IEnKF) in a deterministic framework was\nintroduced in Sakov et al. (2012) to extend the ensemble Kalman filter (EnKF)\nand improve its performance in mildly up to strongly nonlinear cases.\nHowever, the IEnKF assumes that the model is perfect. This assumption\nsimplified the update of the system at a time different from the observation\ntime, which made it natural to apply the IEnKF for smoothing. In this study, we\ngeneralise the IEnKF to the case of imperfect model with additive model error.\nThe new method called IEnKF-Q conducts a Gauss-Newton minimisation in\nensemble space. It combines the propagated analysed ensemble anomalies from the\nprevious cycle and model noise ensemble anomalies into a single ensemble of\nanomalies, and by doing so takes an algebraic form similar to that of the\nIEnKF. The performance of the IEnKF-Q is tested in a number of experiments with\nthe Lorenz-96 model, which show that the method consistently outperforms both\nthe EnKF and the IEnKF naively modified to accommodate additive model noise.\n",
"title": "An iterative ensemble Kalman filter in presence of additive model error"
}
| null | null |
[
"Physics",
"Statistics"
] | null | true | null |
10999
| null |
Validated
| null | null |
null |
{
"abstract": " A statistical algorithm for categorizing different types of matches and fraud\nin image databases is presented. The approach is based on a generative model of\na graph representing images and connections between pairs of identities,\ntrained using properties of a matching algorithm between images.\n",
"title": "A Bayesian algorithm for detecting identity matches and fraud in image databases"
}
| null | null | null | null | true | null |
11000
| null |
Default
| null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.