arXiv subjects

Li Wang

Publications and source records attributed to Li Wang.

At least 469 records · Page 26Linked to original sources

arXiv2019

Towards Ultrahigh Dimensional Feature Selection for Big Data

In this paper, we present a new adaptive feature scaling scheme for ultrahigh-dimensional feature selection on Big Data. To solve this problem effectively, we first reformulate it as a convex semi-infinite programming (SIP) problem and then propose an efficient \emph{feature generating paradigm}. In contrast with traditional gradient-based approaches that conduct optimization on all input features, the proposed method iteratively activates a group of features and solves a sequence of multiple kernel learning (MKL) subproblems of much reduced scale. To further speed up the training, we propose to solve the MKL subproblems in their primal forms through a modified accelerated proximal gradient approach. Due to such an optimization scheme, some efficient cache techniques are also developed. The feature generating paradigm can guarantee that the solution converges globally under mild conditions and achieve lower feature selection bias. Moreover, the proposed method can tackle two challenging tasks in feature selection: 1) group-based feature selection with complex structures and 2) nonlinear feature selection with explicit feature mappings. Comprehensive experiments on a wide range of synthetic and real-world datasets containing tens of million data points with $O(10^{14})$ features demonstrate the competitive performance of the proposed method over state-of-the-art feature selection methods in terms of generalization performance and training efficiency.

cs.LG↗

arXiv2019

Large-Scale Semi-Supervised Learning via Graph Structure Learning over High-Dense Points

We focus on developing a novel scalable graph-based semi-supervised learning (SSL) method for a small number of labeled data and a large amount of unlabeled data. Due to the lack of labeled data and the availability of large-scale unlabeled data, existing SSL methods usually encounter either suboptimal performance because of an improper graph or the high computational complexity of the large-scale optimization problem. In this paper, we propose to address both challenging problems by constructing a proper graph for graph-based SSL methods. Different from existing approaches, we simultaneously learn a small set of vertexes to characterize the high-dense regions of the input data and a graph to depict the relationships among these vertexes. A novel approach is then proposed to construct the graph of the input data from the learned graph of a small number of vertexes with some preferred properties. Without explicitly calculating the constructed graph of inputs, two transductive graph-based SSL approaches are presented with the computational complexity in linear with the number of input data. Extensive experiments on synthetic data and real datasets of varied sizes demonstrate that the proposed method is not only scalable for large-scale data, but also achieve good classification performance, especially for extremely small number of labels.

cs.LG↗

arXiv2019

Electron-positron pair production in combined Sauter potential wells

Electron-positron pair production, in combined Sauter potential wells and an oscillating one is imposed on a static Sauter potential, is investigated by using the computational quantum field theory. We find that the gain number (the difference of pair number under combined potentials to the simple addition of pair number for each potential) of the created pairs depends strongly on the depth of static potential and the frequency of oscillating potential. In particular, it is more sensitive to the frequency compared with the depth. For the low-frequency multiphoton regime, the gaining is almost positive and exhibits interesting nonlinear characteristics on both depth and frequency. For the single-photon regime, however, the gaining is almost negative and decreases near linearly with depth while it exhibits an oscillation characteristic with frequency. Furthermore, the optimal frequency and depth of gain number are found and discussed.

quant-ph↗

arXiv2019

On a compact trace embedding theorem in Musielak-Sobolev spaces

By a stronger compact boundary embedding theorem in Musielak-Orlicz-Sobolev space developed in the paper, variational method is employed to deal with the nonlinear elliptic equation with the nonlinear Neumann boundary condition in the framework of Musielak-Orlicz-Sobolev space.

math.FA↗

arXiv2019

There is Limited Correlation between Coverage and Robustness for Deep Neural Networks

Deep neural networks (DNN) are increasingly applied in safety-critical systems, e.g., for face recognition, autonomous car control and malware detection. It is also shown that DNNs are subject to attacks such as adversarial perturbation and thus must be properly tested. Many coverage criteria for DNN since have been proposed, inspired by the success of code coverage criteria for software programs. The expectation is that if a DNN is a well tested (and retrained) according to such coverage criteria, it is more likely to be robust. In this work, we conduct an empirical study to evaluate the relationship between coverage, robustness and attack/defense metrics for DNN. Our study is the largest to date and systematically done based on 100 DNN models and 25 metrics. One of our findings is that there is limited correlation between coverage and robustness, i.e., improving coverage does not help improve the robustness. Our dataset and implementation have been made available to serve as a benchmark for future studies on testing DNN.

cs.LG↗

arXiv2019

Simultaneous Confidence Band for Stationary Covariance Function of Dense Functional Data

Inference via simultaneous confidence band is studied for stationary covariance function of dense functional data. A two-stage estimation procedure is proposed based on spline approximation, the first stage involving estimation of all the individual trajectories and the second stage involving estimation of the covariance function through smoothing the empirical covariance function. The proposed covariance estimator is smooth and as efficient as the oracle estimator when all individual trajectories are known. An asymptotic simultaneous confidence band (SCB) is developed for the true covariance function, and the coverage probabilities are shown to be asymptotically correct. Simulation experiments are conducted on the numerical performance of the proposed estimator and SCB. The proposed method is also illustrated by two real data examples.

stat.ME↗

arXiv2019

Towards understanding the boundary propagation speeds in tumor growth models

At the continuous level, we consider two types of tumor growth models: the cell density model, which is based on the fluid mechanical construction, is more favorable for scientific interpretation and numerical simulations; and the free boundary model, as the incompressible limit of the former, is more tractable when investigating the boundary propagation. In this work, we aim to investigate the boundary propagation speeds in those models based on asymptotic analysis of the free boundary model and efficient numerical simulations of the cell density model. We derive, for the first time, some analytical solutions for the free boundary model with pressures jumps across the tumor boundary in multi-dimensions with finite tumor sizes. We further show that in the large radius limit, the analytical solutions to the free boundary model in one and multiple spatial dimensions converge to traveling wave solutions. The convergence rate in the propagation speeds are algebraic in multi-dimensions as opposed to the exponential convergence in 1D. We also propose an accurate front capturing numerical scheme for the cell density model, and extensive numerical tests are provided to verify the analytical findings.

math.AP↗

arXiv2019

Unsupervised Multi-stream Highlight detection for the Game "Honor of Kings"

With the increasing popularity of E-sport live, Highlight Flashback has been a critical functionality of live platforms, which aggregates the overall exciting fighting scenes in a few seconds. In this paper, we introduce a novel training strategy without any additional annotation to automatically generate highlights for game video live. Considering that the existing manual edited clips contain more highlights than long game live videos, we perform pair-wise ranking constraints across clips from edited and long live videos. A multi-stream framework is also proposed to fuse spatial, temporal as well as audio features extracted from videos. To evaluate our method, we test on long game live videos with an average length of about 15 minutes. Extensive experimental results on videos demonstrate its satisfying performance on highlights generation and effectiveness by the fusion of three streams.

cs.CV↗

arXiv2019

Hetero-Center Loss for Cross-Modality Person Re-Identification

Cross-modality person re-identification is a challenging problem which retrieves a given pedestrian image in RGB modality among all the gallery images in infrared modality. The task can address the limitation of RGB-based person Re-ID in dark environments. Existing researches mainly focus on enlarging inter-class differences of feature to solve the problem. However, few studies investigate improving intra-class cross-modality similarity, which is important for this issue. In this paper, we propose a novel loss function, called Hetero-Center loss (HC loss) to reduce the intra-class cross-modality variations. Specifically, HC loss can supervise the network learning the cross-modality invariant information by constraining the intra-class center distance between two heterogenous modalities. With the joint supervision of Cross-Entropy (CE) loss and HC loss, the network is trained to achieve two vital objectives, inter-class discrepancy and intra-class cross-modality similarity as much as possible. Besides, we propose a simple and high-performance network architecture to learn local feature representations for cross-modality person re-identification, which can be a baseline for future research. Extensive experiments indicate the effectiveness of the proposed methods, which outperform state-of-the-art methods by a wide margin.

cs.CV↗

arXiv2019

Characterizing Membership Privacy in Stochastic Gradient Langevin Dynamics

Bayesian deep learning is recently regarded as an intrinsic way to characterize the weight uncertainty of deep neural networks~(DNNs). Stochastic Gradient Langevin Dynamics~(SGLD) is an effective method to enable Bayesian deep learning on large-scale datasets. Previous theoretical studies have shown various appealing properties of SGLD, ranging from the convergence properties to the generalization bounds. In this paper, we study the properties of SGLD from a novel perspective of membership privacy protection (i.e., preventing the membership attack). The membership attack, which aims to determine whether a specific sample is used for training a given DNN model, has emerged as a common threat against deep learning algorithms. To this end, we build a theoretical framework to analyze the information leakage (w.r.t. the training dataset) of a model trained using SGLD. Based on this framework, we demonstrate that SGLD can prevent the information leakage of the training dataset to a certain extent. Moreover, our theoretical analysis can be naturally extended to other types of Stochastic Gradient Markov Chain Monte Carlo (SG-MCMC) methods. Empirical results on different datasets and models verify our theoretical findings and suggest that the SGLD algorithm can not only reduce the information leakage but also improve the generalization ability of the DNN models in real-world applications.

cs.LG↗

arXiv2019

Generalization in Generative Adversarial Networks: A Novel Perspective from Privacy Protection

In this paper, we aim to understand the generalization properties of generative adversarial networks (GANs) from a new perspective of privacy protection. Theoretically, we prove that a differentially private learning algorithm used for training the GAN does not overfit to a certain degree, i.e., the generalization gap can be bounded. Moreover, some recent works, such as the Bayesian GAN, can be re-interpreted based on our theoretical insight from privacy protection. Quantitatively, to evaluate the information leakage of well-trained GAN models, we perform various membership attacks on these models. The results show that previous Lipschitz regularization techniques are effective in not only reducing the generalization gap but also alleviating the information leakage of the training dataset.

cs.LG↗

arXiv2019

A Self-consistent-field Iteration for Orthogonal Canonical Correlation Analysis

We propose an efficient algorithm for solving orthogonal canonical correlation analysis (OCCA) in the form of trace-fractional structure and orthogonal linear projections. Even though orthogonality has been widely used and proved to be a useful criterion for pattern recognition and feature extraction, existing methods for solving OCCA problem are either numerical unstable by relying on a deflation scheme, or less efficient by directly using generic optimization methods. In this paper, we propose an alternating numerical scheme whose core is the sub-maximization problem in the trace-fractional form with an orthogonal constraint. A customized self-consistent-field (SCF) iteration for this sub-maximization problem is devised. It is proved that the SCF iteration is globally convergent to a KKT point and that the alternating numerical scheme always converges. We further formulate a new trace-fractional maximization problem for orthogonal multiset CCA (OMCCA) and then propose an efficient algorithm with an either Jacobi-style or Gauss-Seidel-style updating scheme based on the same SCF iteration. Extensive experiments are conducted to evaluate the proposed algorithms against existing methods including two real world applications: multi-label classification and multi-view feature extraction. Experimental results show that our methods not only perform competitively to or better than baselines but also are more efficient.

cs.LG↗

arXiv2019

Core-Shell Nanofiber Containing Large Amount of Flame Retardants via Coaxial Dual-Nozzle Electrospinning as Battery Separators

Lithium-ion batteries have attracted enormous interests recently as promising power sources. However, the safety issue associated with the employment of highly flammable liquid electrolyte impedes the further development of next-generation lithium-ion batteries. Recently, researchers reported the use of electrospun core-shell fiber as the battery separator consisting of polymer layer as protective shell and flame retardants loaded inside as core. In case of a typical battery shorting, the protective polymer shell melts during thermal-runaway and the flame retardants inside would be released to suppress the combustion of the electrolyte. Due to the use of a single precursor solution for electrospinning containing both polymer and flame retardants, the weight ratio of flame retardants is limited and dependent. Herein, we developed a dual-nozzle, coaxial electrospinning approach to fabricate the core-shell nanofiber with a greatly enhanced flame retardants weight percentage in the final fibers. The weight ratio of flame retardants of triphenyl phosphate in the final composite reaches over 60 wt.%. The LiFePO4-based cell using this composite nanofiber as battery separator exhibits excellent flame-retardant property without compromising the cycling stability or rate performances. In addition, this functional nanofiber can also be coated onto commercial separators instead of being used directly as separators.

physics.app-ph↗

arXiv2019

Accurate front capturing asymptotic preserving scheme for nonlinear gray radiative transfer equation

We develop an asymptotic preserving scheme for the gray radiative transfer equation. Two asymptotic regimes are considered: one is a diffusive regime described by a nonlinear diffusion equation for the material temperature; the other is a free streaming regime with zero opacity. To alleviate the restriction on time step and capture the correct front propagation in the diffusion limit, an implicit treatment is crucial. However, this often involves a large-scale nonlinear iterative solver as the spatial and angular dimensions are coupled. Our idea is to introduce an auxiliary variable that leads to a ``redundant" system, which is then solved with a three-stage update: prediction, correction, and projection. The benefit of this approach is that the implicit system is local to each spatial element, independent of angular variable, and thus only requires a scalar Newton's solver. We also introduce a spatial discretization with a compact stencil based on even-odd decomposition. Our method preserves both the nonlinear diffusion limit with correct front propagation speed and the free streaming limit, with a hyperbolic CFL condition.

math.NA↗

arXiv2019

Barrier-Certified Adaptive Reinforcement Learning with Applications to Brushbot Navigation

This paper presents a safe learning framework that employs an adaptive model learning algorithm together with barrier certificates for systems with possibly nonstationary agent dynamics. To extract the dynamic structure of the model, we use a sparse optimization technique. We use the learned model in combination with control barrier certificates which constrain policies (feedback controllers) in order to maintain safety, which refers to avoiding particular undesirable regions of the state space. Under certain conditions, recovery of safety in the sense of Lyapunov stability after violations of safety due to the nonstationarity is guaranteed. In addition, we reformulate an action-value function approximation to make any kernel-based nonlinear function estimation method applicable to our adaptive learning framework. Lastly, solutions to the barrier-certified policy optimization are guaranteed to be globally optimal, ensuring the greedy policy improvement under mild conditions. The resulting framework is validated via simulations of a quadrotor, which has previously been used under stationarity assumptions in the safe learnings literature, and is then tested on a real robot, the brushbot, whose dynamics is unknown, highly complex and nonstationary.

cs.LG↗

arXiv2019

Fisher information regularization schemes for Wasserstein gradient flows

We propose a variational scheme for computing Wasserstein gradient flows. The scheme builds upon the Jordan--Kinderlehrer--Otto framework with the Benamou-Brenier's dynamic formulation of the quadratic Wasserstein metric and adds a regularization by the Fisher information. This regularization can be derived in terms of energy splitting and is closely related to the Schr{ö}dinger bridge problem. It improves the convexity of the variational problem and automatically preserves the non-negativity of the solution. As a result, it allows us to apply sequential quadratic programming to solve the sub-optimization problem. We further save the computational cost by showing that no additional time interpolation is needed in the underlying dynamic formulation of the Wasserstein-2 metric, and therefore, the dimension of the problem is vastly reduced. Several numerical examples, including porous media equation, nonlinear Fokker-Planck equation, aggregation diffusion equation, and Derrida-Lebowitz-Speer-Spohn equation, are provided. These examples demonstrate the simplicity and stableness of the proposed scheme.

math.NA↗

arXiv2019

Hierarchical Deep Multiagent Reinforcement Learning with Temporal Abstraction

Multiagent reinforcement learning (MARL) is commonly considered to suffer from non-stationary environments and exponentially increasing policy space. It would be even more challenging when rewards are sparse and delayed over long trajectories. In this paper, we study hierarchical deep MARL in cooperative multiagent problems with sparse and delayed reward. With temporal abstraction, we decompose the problem into a hierarchy of different time scales and investigate how agents can learn high-level coordination based on the independent skills learned at the low level. Three hierarchical deep MARL architectures are proposed to learn hierarchical policies under different MARL paradigms. Besides, we propose a new experience replay mechanism to alleviate the issue of the sparse transitions at the high level of abstraction and the non-stationarity of multiagent learning. We empirically demonstrate the effectiveness of our approaches in two domains with extremely sparse feedback: (1) a variety of Multiagent Trash Collection tasks, and (2) a challenging online mobile game, i.e., Fever Basketball Defense.

cs.LG↗

arXiv2019

Probabilistic Structure Learning for EEG/MEG Source Imaging with Hierarchical Graph Prior

Brain source imaging is an important method for noninvasively characterizing brain activity using Electroencephalogram (EEG) or Magnetoencephalography (MEG) recordings. Traditional EEG/MEG Source Imaging (ESI) methods usually assume that either source activity at different time points is unrelated, or that similar spatiotemporal patterns exist across an entire study period. The former assumption makes ESI analyses sensitive to noise, while the latter renders ESI analyses unable to account for time-varying patterns of activity. To effectively deal with noise while maintaining flexibility and continuity among brain activation patterns, we propose a novel probabilistic ESI model based on a hierarchical graph prior. Under our method, a spanning tree constraint ensures that activity patterns have spatiotemporal continuity. An efficient algorithm based on alternating convex search is presented to solve the proposed model and is provably convergent. Comprehensive numerical studies using synthetic data on a real brain model are conducted under different levels of signal-to-noise ratio (SNR) from both sensor and source spaces. We also examine the EEG/MEG data in a real application, in which our ESI reconstructions are neurologically plausible. All the results demonstrate significant improvements of the proposed algorithm over the benchmark methods in terms of source localization performance, especially at high noise levels.

stat.AP↗