Profile

I am a Ph.D. student majoring in Computer Science at Shanghai Jiao Tong University, a member of the Wen-Tsun Wu AI honorary doctoral class. I obtained my B.Eng. majoring in CS at SJTU as a member of the ACM Honors Class, an elite CS program for the top 5% of talented students.

Experiences

Ph.D. Researcher

2020.9 - present
Artificial Intelligence Institute, Shanghai Jiao Tong University
  • I am a Ph.D. researcher, advised by Prof. Hongyuan Zha and Prof. Junchi Yan.
  • Focus on utilizing machine learning methods to solve combinatorial optimization problems in the real world.

Research Intern

2022.1 - 2022.1
Decision Making and Reasoning Laboratory, Huawei Noah’s Ark Laboratory
  • I am a research intern, advised by Prof. Xijun Li and Prof. Jia Zeng.
  • Focus on using deep learning approaches to improve the solving of mix-integer programming (MIP).
  • Propose to improve the presolving part in model MIP solvers. The results on both common MIP datasets and Huawei’s industrial scenarios have demonstrated the performance of our proposed method.

Undergraduate Researcher

2020.2 - 2020.6
APEX Data & Knowledge Management Laboratory, Shanghai Jiao Tong University
  • I am an undergraduate researcher, advised by Prof. Weinan Zhang.
  • Focus on improving the generalization ability of traffic signal control models.
  • Propose a novel meta-RL framework and conduct extensive experiments.

Research Intern

2019.8 - 2020.1
Pennsylvania State University
  • I am a research intern, advised by Prof. Jessie Li.
  • Focus on “learning to simulate”, making the simulator closer to the real world.
  • Try to use the combination of physic model and deep neural networks, for a good performance when the data is insufficient.

Undergraduate Researcher

2018.7 - 2019.7
Data Driven Software Technology Laboratory, Shanghai Jiao Tong University
  • I am an undergraduate researcher, advised by Prof. Yanmin Zhu.
  • Focus on urban computing and urban data mining.
  • Conduct a comprehensive study on exploring.

Publications

  • Towards General Loop Invariant Generation via Coordinating Symbolic Execution and Large Language Model
  • Chang Liu, Xiwei Wu, Yuan Feng, Qinxiang Cao, Junchi Yan.
    NeurIPS, 2024
  • L2P-MIP Learning to Presolve for Mixed Integer Programming
  • Chang Liu, Zhichen Dong, Haobo Ma, Weilin Luo, Xijun Li, Bowen Pang, Jia Zeng, Junchi Yan.
    ICLR, 2024
  • Revocable Deep Reinforcement Learning with Affinity Regularization for Outlier-Robust Graph Matching
  • Chang Liu, Zetian Jiang, Runzhong Wang, Lingxiao Huang, Pinyan Lu, Junchi Yan.
    ICLR, 2023
  • A Survey for Solving Mixed Integer Programming via Machine Learning
  • Jiayi Zhang, Chang Liu, Xijun Li, Hui-Ling Zhen, Mingxuan Yuan, Yawen Li, Junchi Yan.
    Neurocomputing
  • Self-supervised Learning of Visual Graph Matching
  • Chang Liu*(equal contribution), Shaofeng Zhang*, Xiaokang Yang, Junchi Yan.
    ECCV, 2022
  • Deep Neural Network Fusion via Graph Matching with Applications to Model Ensemble and Federated Learning
  • Chang Liu, Chenfei Lou, Runzhong Wang, Yuhan Xi, Li Shen, Junchi Yan.
    ICML, 2022
  • Knowledge-based Residual Learning
  • Guanjie Zheng, Chang Liu, Hua Wei, Porter Jenkins, Chacha Chen, Tao Wen, Zhenhui Li.
    IJCAI, 2021
  • Rebuilding City-Wide Traffic from Multi-Source Data
  • Chang Liu*(equal contribution), Guanjie Zheng*, Hua Wei, Chacha Chen, Zhenhui Li.
    ICDE, 2021
  • GeneraLight Improving Environment Generalization of Traffic Signal Control via Meta Reinforcement Learning
  • Chang Liu*(equal contribution), Huichu Zhang*, Weinan Zhang, Guanjie Zheng, Yong Yu
    CIKM, 2020
  • Learning to Simulate on Sparse Trajectory Data (Best ADS Paper Award)
  • Hua Wei, Chacha Chen, Chang Liu, Guanjie Zheng, Zhenhui Li
    ECML-PKDD, 2020
  • CityFlow A Multi-Agent Reinforcement Learning Environment for Large Scale City Traffic Scenario
  • Huichu Zhang, Siyuan Feng, Chang Liu, Yaoyao Ding, Yichen Zhu, Zihan Zhou, Weinan Zhang, Yong Yu, Haiming Jin, Zhenhui Li
    WWW, 2019

    Projects

    Machine Learning for Combinatorial Optimization Problems - We maintain a list of resources that utilize machine learning technologies to solve combinatorial optimization problems (ML4CO), which has gained more than 1,000 stars on Github.
    Generalized Traffic Signal Control System - We propose a novel meta-RL framework GeneraLight for traffic light signal control models, which can significantly improve their generalization ability to different environments.
    Learning to Simulate - We focus on reducing the gap between simulator and the real world, finding solutions to different problems.
    City Traffic Simulator - We Built a city traffic simulator, serving reinforcement learning algorithm as a multi-agent reinforcement learning environment.
    Shared Bike System in Urban Computing - Focusing on transfer learning and spatial-temporal analysis in shared bike system, I wrote a paper proposing a model combined with principle component analysis, discrete wavelet transform and attention mechanism.
    Mx Compiler - I designed a compiler implemented in Java from scratch, translating a Java-and-C-like language to NASM, and implemented optimizations for the compiler, faster than gcc O1 on some test sets.
    Shor Algorithm - A simple implement of shor algorithm in quantum computing, written by a new quantum language called Q#.