Chengkai Han
Student at Beihang University(BUAA)
Beijing, China
Summary
🤩
RockstarChengkai Han is a graduate student at Beihang University's School of Computer Science and Engineering with six years of hands-on experience in machine learning and spatio-temporal modeling. Based in Beijing, he contributes to open-source urban data projects like LibCity, where he implemented and refined TGCLSTM, ResLSTM and STDN modules—demonstrating strength in model architecture, dataset integration, and low-level module design such as a FilterLinear. His work bridges research and production-ready code, focusing on temporal graph convolutional LSTM variants for city-scale prediction tasks. Notably, he often contributes to core model functionality rather than peripheral tooling, indicating a deep comfort with algorithmic detail and implementation-level fixes.
6 years of coding experience
Github Skills (6)
pytorch10
deeplearning-ai10
deep-learning10
n9
rnn-model9
lstm9
Programming languages (4)
Github contributions (5)
LibCity/Bigscity-LibCity
Feb 2021 - May 2022
LibCity: An Open Library for Urban Spatial-temporal Data Mining
Role in this project:
ML Engineer
ML EngineerContributions:33 reviews, 19 commits, 23 PRs in 1 year 2 months
Contributions summary:Chengkai contributed to the development of a TGCLSTM model, specifically adding, modifying, and fixing the implementation of this spatio-temporal prediction model. The commits show the creation of a FilterLinear module and adjustments to the TGCLSTM class, indicating a focus on model architecture and functionality. Furthermore, the user worked on incorporating a new STDN model and a ResLSTM model and providing changes in the dataset used for each one. These changes demonstrate the user's involvement in the core machine learning aspects of the project.
estimated-time-of-arrivaletarepresentation-learningdata-miningmap-matching
NickHan-cs/Spatio-Temporal-Data-Mining-Survey
Jul 2021 - Sep 2022
Contributions:17 commits, 22 pushes, 1 branch in 1 year 2 months
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.