A5032279163- Sparse and Compressive Sensing Techniques
- Photoacoustic and Ultrasonic Imaging
- Software-Defined Networks and 5G
- Structural Health Monitoring Techniques
- Machine Fault Diagnosis Techniques
- Prostate Cancer Diagnosis and Treatment
- Spectroscopy Techniques in Biomedical and Chemical Research
- Advanced Optical Network Technologies
- Gear and Bearing Dynamics Analysis
- Numerical methods in inverse problems
- Statistical and numerical algorithms
- Fault Detection and Control Systems
- Advanced Wireless Network Optimization
- Image and Video Quality Assessment
- Video Coding and Compression Technologies
- Multimedia Communication and Technology
- Optimization and Search Problems
- Spectroscopy and Chemometric Analyses
- Advanced Measurement and Detection Methods
- Network Security and Intrusion Detection
- Satellite Communication Systems
- Image and Signal Denoising Methods
- Statistical Methods and Inference
- Advanced Photonic Communication Systems
- Advanced Radiotherapy Techniques
Memorial Sloan Kettering Cancer Center
2024
New York University
2018-2022
Beijing University of Posts and Telecommunications
2020-2022
ORCID
2021
Vibration monitoring is one of the most effective ways for bearing fault diagnosis, and a challenge how to accurately estimate signals from noisy vibration signals. In this paper, nonconvex sparse regularization method diagnosis proposed based on generalized minimax-concave (GMC) penalty, which maintains convexity sparsity-regularized least squares cost function, thus global minimum can be solved by convex optimization algorithms. Furthermore, we introduce k-sparsity strategy adaptive...
Non-convex sparsity-inducing penalties outperform their convex counterparts, but generally sacrifice the cost function convexity. As a middle ground, we propose sharpening sparse regularizers (SSR) framework to design non-separable non-convex that induce sparsity more effectively than such as <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">l</i> <sub xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> and nuclear norms, without sacrificing The...
Energy consumption is becoming a key issue in the research of future network. In practice, network traffic has periodic time distribution that occurs most often at low level. This feature provides possibility achieving energy savings through topology switching. By considering deficiencies existing studies, such as adaptability between working and load, abnormal switching caused by unbalanced traffic, reliability energy-saving topology, this paper proposes an energy-efficient routing method...
For the suppression of transient artifacts in time series data, we propose a non-convex generalized fused lasso penalty for estimation signals comprising low-pass signal, sparse piecewise constant and additive white Gaussian noise. The proposed is designed so as to preserve convexity total cost function be minimized, thereby realizing benefits convex optimization framework (reliable, robust algorithms, etc.). Compared conventional use L1 norm penalty, does not underestimate true amplitude...
The adaptive bitrate (ABR) algorithm based on reinforcement learning (RL) can actively learn control policies and adapt to different network environments. However, the user quality of experience (QoE) as optimization objective contains multiple indicators. In case complex task objective, whether reward feedback from environment effectively guide policy update is challenge RL-based ABR algorithm. We propose an intrinsic reward, which encourages exploration by enhancing agent's motivation...
Fused lasso norm is classically adopted to model sparse piecewise constant signals, however it not the convex hull of best representation such simultaneously structured signal. In this paper, we propose a variational for better modeling signals. The based on (1) promoting sparsity in first-order difference with total variation and (2) exploiting latent group structure simple linear constraints. We demonstrate proposed outperforms fused denoising setup numerical experiments.
The demand for satellite relay service is increasing, while the network resources are limited and unevenly distributed, which pose a great challenge to task scheduling of tracking data satellites. Most existing models based on static antenna setup time, has limitations in practical applications leads ineffective utilization resources. This paper problem dynamic time splittable tasks, maximizes total scheduled number minimizes time. We also propose two-stage insertion heuristic solve problem....