Heng Chang
Visiting PHD Student at Tencent
Haidian District, Beijing, China
Summary
🤩
Rockstar🎓
Top SchoolHeng Chang is a visiting PhD student and machine learning engineer with nine years of experience bridging academic research and industry R&D across Tsinghua, UC Berkeley, Tencent, and WorldQuant. His doctoral work in data science and prior double major in engineering and management reflect a rare mix of deep technical rigor and business sensibility. He contributes to open-source graph ML tooling (notably enhancements to the AutoGL framework supporting DGL, GIN, and HAN models), and has practical experience fixing real-world data/configuration and parallel-processing issues. At Berkeley's Prof. Sojoudi lab and Tencent AI Lab he has focused on scalable ML systems and social propagation research, applying both quantitative and engineering skills to production-style codebases. Colleagues value him for quickly turning research ideas into robust, reproducible software.
9 years of coding experience
2 years of employment as a software developer
电子工程系电子信息类, 电子工程系电子信息类 at 清华大学Management (Double Major), Management (Double Major) at 清华大学经济管理学院 Tsinghua University School of Economics and Management
Spanish, English, Chinese
Github Skills (6)
pytorch10
machine-learning10
deep-learning10
graph-neural-network10
automl9
dgl9
Programming languages (2)
Github contributions (5)
THUMNLab/AutoGL
Dec 2020 - Apr 2022
An autoML framework & toolkit for machine learning on graphs.
Role in this project:
ML Engineer
ML EngineerContributions:33 commits, 3 PRs, 22 pushes in 1 year 3 months
Contributions summary:Heng primarily focused on improving the `autogl` framework for machine learning on graphs. Their contributions involved fixing dataset and configuration issues, specifically addressing problems related to masks and configurations within the GradientBoostingClassifier. The user also made changes to the training process and data loading utilities, optimizing the use of multiple workers and setting the start method for parallel processing. Furthermore, they added examples for the DGL backend, and improved the supporting of GIN/HAN models within the framework.
pytorchdata-sciencedeep-learningneural-architecture-searchhyper-parameter-optimization
SwiftieH/IGNN
Oct 2020 - Nov 2021
Contributions:8 commits, 6 pushes, 1 branch in 1 year 1 month
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.