Area of Interest |
Scalable Machine Learning, Cloud/Edge Intelligence, Optimization, Convergence Analysis, Federated Learning and Differential Privacy.
|
Education |
Ph.D. Purdue University, USA Aug. 2019 to Aug. 2023
- Major: Electrical and Computer Engineering (ECE) - Thesis: Distributed Machine Learning over Large-scale Networks (Advisor: Prof. Nicolò Michelusi and Prof. Christopher Brinton) M.S. National Taiwan University (NTU), Taiwan Sept. 2016 to June 2018 - Major: Electrical Engineering (Communication Engineering) - Thesis: A Hierarchical Edge-Cloud SDN Controller System with Optimal Resource Allocations for Load-Balancing (Advisor: Prof. Zsehong Tsai) B.S. National Sun Yat-sen University (NSYSU), Taiwan Sept. 2012 to June 2016 - Major: Electrical Engineering - Independent Study: Approaches with Sparse Perfect Gaussian Integer Sequences for Physical Layer Security in Wireless Network (Advisor: Prof. Chih-Peng Li) |
Honor & Distinctions
|
|
Leadership Experiences
|
|
Professional Society Activities |
Reviewer for the following Journals and Conference proceedings
- IEEE Transactions on Signal Processing - IEEE Transactions on Mobile Computing - IEEE Journal on Selected Areas in Communications - IEEE Transactions on Wireless Communications - IEEE Transactions on Communications - IEEE Communications Magazine - IEEE Access - IEEE International Conference on Computer Communications (IEEE INFOCOM) - IEEE Global Communications Conference (IEEE Globecom) |
Research Experience |
School of Electrical and computer engineering at Purdue University. 2019 to current
- Design of scalable machine learning algorithms: Design learning algorithms based on mathematical analysis to efficiently solve a general class of machine learning problems over large-scale networks. ([J1], [C1], [C2]) - Characterizing the design of learning parameters: Theoretical analysis on the the learning algorithm to reveal the optimal decision of tunable learning parameters and its effect on the learning performance. ([J1], [C1], [C2]) (Advisor: Dr. Christopher Brinton, Dr. Nicolò Michelusi) Institute of Communications Engineering in NTU. 2016 to 2018 - Architecture Design for Large-Scale SDN Structure: Developed a Hierarchical SDN system with a new load-balancing algorithm. Simulation results shows that the system provides outstanding results in both delay and fairness in large-scale network. ([J2]) - Queueing Model Design for SDN: Developed a simulation model for Hierarchical SDN systems using queueing theory. Mathematical models can be efficiently derived according to the requirements of the network. ([J2]) - Evolutionary Algorithm Design for Community Detection: Developed a swarm intelligence-based community detection algorithm. Computation Complexity is highly reduced and different scales of communities can be revealed through designs of particle update in each iteration. (Advisor: Dr. Zsehong Tsai) Institute of Statistical Science in Academia Sinica. 2015 to 2018 - Scheduling Design for Supercomputers: Propose and implement an effective learning and scheduling algorithm design on the ALPS supercomputer. Simulation results on various job arrivals show that our algorithm outperforms the state-of-the-art approaches with at least 100-times speedup. ([J3]) - Trajectory Localization: Design a trajectory localization evolutionary algorithm to construct the confidence regions of target trajectory. The moving speed of the trajectory can be easily inferred from the state of the confidence regions without additional hardware. ([J4]) - Parallel Design on Evolutionary Algorithms: Developed a parallizable structure for swarm intelligence-based evolutionary algorithm. The proposed work offers a good approach for parallel evolutionary computation on CPU. ([C3]) (Advisor: Dr. Frederick Kin Hing Phoa) Department of Electrical Engineering in NSYSU 2015 to 2016 - Wireless Security Design for Eavesdroppers in OFDM systems: Developed a low-complexity cryptography encryption method and a simple four-phase channel estimation scheme to provide secure channel. Simulation results demonstrates that the eavesdropper can not obtain any information about transmitted data even located near a legitimate transmitter. ([C4]) (Advisor: Dr. Chih-Peng Li) |
Working Experiences |
|
Mentor |
Dr. Jacob C. Huang
|