Graduate Research Assistant, Sensor Fusion Lab, Syracuse University, NY Aug. 2014 - present
Advisor: Prof. Pramod K. Varshney
Project ongoing: Fundamental understandings of generative adversarial networks (GANs)
Project 1: Algorithm Design of Accelerated Proximal Methods for Non-convex and Stochastic Optimization
Project description
- Proposed a simple accelerated proximal method that can both work in convex and nonconvex cases.
- Established convergence in nonconvex case for accelerated proximal SVRG with the KL condition.
- With inexact proximal mapping and subgradient calculating, established convergence results in both convex and nonconvex cases.
Project 2: Influential Node Detection in Implicit Social Networks
Project description
- Modeled the information diffusion of multiple contagions in implicit social networks.
- Proposed influence characterization with sparsity and low rank constraints for influential node detection.
- Exploited contagion correlation by copula theory to improve the regression performance with collected ISIS tweets data set.
- Formulated an optimization problem with multiple regularizers, and solved it using graphical Lasso based inverse variance estimation, incremental proximal descent, and alternating approach.
Project 3: Aggregation Rule Design for Reliable Labeling by Crowdsourcing
Project description
- Proposed a general framework of weight assignment to crowd’s response for aggregation and fusion.
- Established the optimal aggregation rules for reliable labeling by crowdsourcing with or without spammers.
- Developed effective schemes for reliable labeling when the spammers behave strategically.
Research Internship, Lawrence Livermore National Laboratory, Livermore, CA May - Aug 2017
Division : Computational Engineering
Project : Collaborative Autonomy
Project description
- Established convergence guarantees for decentralized ADMM with attacks and errors
- Developed a general framework to evaluate the effectiveness of robust algorithms for collaborative autonomy
- Proposed an effective decentralized ADMM for robust collaborative autonomy
Graduate Research Assistant, State Key Lab of Integrated Service Networks, Xidian University, Xi'an, China
Project1: Resource Management in High-Throughput Cognitive Radio Networks
Project description
- 1. Designed an adaptive two-way cognitive relay network with relay selection and power allocation.
- 2. Proposed a framework for beamforming and power splitting to guarantee secure communication and energy harvesting.
- 3. Developed algorithms for convex and nonconvex optimization problems with practical constraints in radio networks.
Project 2: Stochastic Signal Enhancement by Artificially Adding Noise
Project description
- 1. Designed efficient signal enhancement techniques exploiting resonance effects between signal and noise.
- 2. Proposed a highly applicable nonlinear framework for signal enhancement by adding artificial noise.
- 3. Established theoretical guarantees for signal enhancement with artificial noise.