Artificial Intelligence undergraduate at Beijing Normal University–Hong Kong Baptist University United International College (BNBU). I build reproducible machine-learning experiments and end-to-end software systems, with particular interest in interpretable AI, reliable evaluation, and AI applications.
Python · Java · JavaScript · Machine Learning · Interpretable AI · Software Engineering · Full-stack Development
-
Auditing an Interpretable Weather–Symptom–Disease Cascade
A leakage-resistant audit of a concept-bottleneck pipeline, with grouped evaluation, robustness checks, a controlled validation study, and explicit non-clinical research boundaries. -
BNBU BookCycle
A WeChat Mini Program for course-aware campus textbook exchange, with two-sided matching, local handoff workflows, and a runnable no-backend demonstration mode. -
CineFlow
A transaction-aware Java desktop cinema ticketing system with JDBC, authorization, row locking, auditability, and a polished Swing interface. -
AI Animal Chess
A Pygame implementation of Dou Shou Qi featuring Negamax, Monte Carlo Tree Search, reinforcement-learning agents, benchmarking, and an in-game AI thinking panel. -
Credit Risk Ensemble Learning From Scratch
An educational tabular machine-learning project with transparent from-scratch ensemble components and reproducible preprocessing.
- TDesign Miniprogram · PR #4580
Improved the
Fabauto-collapse demo for Issue #2709, including behavior tests and documentation. Currently awaiting community review.
I am developing stronger research practice around interpretable machine learning: defining claims carefully, comparing direct and concept-mediated pipelines fairly, and making experimental evidence reviewable.