Education
- B.S. Electronic Engineering, Xidian University, 2021
Competition Awards
- 2018 ACM-ICPC China Invitational Contest Silver Medal
Work&Research experience
2022.11-2023.9: Research Assistant- Microsoft - Microsoft Research Asia - Machine Learning Group
- Contribution: During my time at Microsoft Research Asia, I was solely responsible for a research project. The purpose of this project was to use reinforcement learning and graph neural networks to solve a complex combinatorial optimization problem - the Flexible Job Shop Scheduling Problem (FJSP). I built a simulator to simulate the process of solving this problem and proposed a method that combines generative (Learning to Generate) and improvement (Learning to Improve) approaches to achieve better results than other methods in the literature. This work is currently being submitted to the ICLR2024, and the code for the related work has been uploaded to my Github homepage.
- Mentor: Dr. Song Lei,Dr. Weidong Ma
2022.9-2022.11: Research Intern- BaiDu Research - PARL Group
- Contribution: During my time at Baidu, I am responsible for participating in the development and maintenance of PARL, an open source reinforcement learning framework.
- Mentor:Bo Zhou
2022.4-2022.9: Research Assistant- Aviation Industry Corporation of China (AVIC) - China Aviation Research
- Contribution: Designed and developed a reinforcement learning-based stall recovery control system for large aircraft (Boeing 737-800).
Designed and developed a reinforcement learning-based spin recovery control system for fighter jets (F-16).
During the project, two optimization methods were proposed to make the model performance surpass traditional control methods:
-Two-stage reward function model fusion
-Action value exponential filtering - Mentor: Prof. Junqiu Wang
2021.4-2021.7: Algorithm Intern- Hikvision Research Institute-Pre Research Group
- Contribution: During my internship, I built a reinforcement learning-based robotic arm control system. The purpose of this system is to use a deep reinforcement learning network to control a seven-degree-of freedom robotic arm to automatically search for negative class samples required by the target model.
- Mentor: Ying Bian
Skills
- Python
- C++
- Matlab
- JavaScripts
