About
Hsing-Yu Chen 陳幸妤
Ph.D. student at NYCU (CGI Lab, advised by Prof. I-Chen Wu) and research student at Academia Sinica RLG Lab. I work on reinforcement learning and game AI for imperfect-information games — Taiwanese Mahjong, Werewolf, Wall Go — and on HPC and generative-AI systems.
Education
National Yang Ming Chiao Tung University (NYCU) — Hsinchu, Taiwan
- Ph.D., Institute of Computer Science and Engineering · Sep. 2025 – present
- Research focus: Deep Reinforcement Learning, Game AI, Generative AI
National Taichung University of Education (NTCU) — Taichung, Taiwan
- B.S., Department of Computer Science · Sep. 2021 – Jun. 2025
- GPA 4.25 / 4.3 (average 94.08) · Rank 1 / 46 (Top 2.1%)
- Awards: NSTC College Student Research Creativity Award (大專學生研究創作獎), Academic Excellence Awards (菁英獎、智育獎、校友總會優秀獎學金, multiple semesters)
Research Interests
- Deep Reinforcement Learning (DRL) — PG, AlphaZero, MuZero frameworks
- Game AI — imperfect-information games (Mahjong, Werewolf, Wall Go), theory of mind
- High-Performance Computing (HPC) — parallel computing, performance tuning
- Generative AI — LLM applications in education, automated software engineering
Publications
- Chen, H. Y., Arjonilla, J., Wu, I. C., & Wu, T. R. (2026). WallZero: Mastering the Game of WallGo with Strategic Analysis. International Conference on Computers and Games (CG 2026), Maastricht, the Netherlands, June 2026.
- Chen, H. Y., & Huang, K. C. (2025). Improved Opponent Modeling in DRL-Based Taiwanese Mahjong AI. Taiwan Computer Game Association (TCGA 2025), Taiwan, May 2025. — Best Paper Award
- Chang, Y. E., Hung, Y. H., Chen, H. Y., …, & Huang, K. C. (2024). A Sustainable Online Learning Platform for After-Class Peer Learning. eLearn 2024, Singapore, October 7–10, 2024. Paper ID: 63981.
Honors & Awards
International
- Silver Medal — ICGA Computer Olympiad 2024 (Mahjong: Dartrix)
- Bronze Medal — ICGA Computer Olympiad 2024 (Mahjong: Rowlet)
- Qualifier (National Team) — PMI Region 9 Project Management Case Competition 2024
National
- Champion (1st Place) — 2026 TCGA Computer Game Tournament (Mahjong)
- Champion (1st Place) — 2025 TAAI Computer Game Tournament (Mahjong)
- Champion (1st Place) — 2024 TCGA Computer Game Tournament (Mahjong)
- Champion (1st Place) — 3rd NCHC Cup: HPC Application Optimization Competition (國網盃)
- Champion (1st Place) — FIRST LEGO League Taiwan 2018–2019
- Runner-up (2nd Place) — 2025 TCGA Computer Game Tournament (Mahjong)
- Runner-up (2nd Place) — 2024 Taipei City Data Hackathon (GenAI & Microservices)
- 3rd Place (Regional) — Web Design, 54th National Skills Competition
- 5th Place (Regional) — Cloud Computing, 55th National Skills Competition
- National Outstanding Young College Elite (全國大專優秀青年), 2025 — sole representative from ~5,100 NTCU students; one of ~150 national recipients.
- NSTC College Student Research Creativity Award (大專學生研究創作獎), 2024 — ranked in the top 2% of 9,505 research grantees nationwide.
Academic Service
- Reviewer — IEEE Transactions on Games (2025)
- Reviewer — ICGA Journal (2026)
- Subreviewer / External Reviewer — IEEE Conference on Games (CoG 2026)
Work Experience
Developer — NYCU Information Technology Service Center · Sep. 2025 – present
Research Student — Academia Sinica, RLG Lab · Jul. 2025 – present
Research Intern — National Center for High-performance Computing (NCHC), Hsinchu · Mar. 2025 – Jul. 2025
- Investigated advanced LLM techniques — prompt engineering, RAG, and step-by-step distillation — and designed data pipelines for large-scale model training.
Lead Web Developer — Dept. of Computer Science, NTCU · 2022 – 2025
Summer Intern — Shanghai Development Center of Computer Software Technology, Shanghai, China · Jul. 2023 – Aug. 2023
- Web development with Vue.js and PHP (Laravel), plus DevOps.
Research & Project Experience
Deep Reinforcement Learning for Imperfect-Information Games (Taiwanese Mahjong) Principal Investigator — NSTC Grant No. 113-2813-C-142-003-E · Jul. 2023 – Feb. 2025
- Secured the research grant (~45% acceptance rate from a pool of 19,900+ applicants).
- Achievement: optimized MCTS significantly improved win rates against the baseline (Rowlet) — Silver Medal at ICGA 2024, Best Paper Award at TCGA 2025, and NSTC College Student Research Creativity Award (top 2% of 9,505 projects).
- Problem: traditional MCTS fails in high-stochasticity games like Mahjong due to hidden information.
- Methodology: developed Dartrix, a Policy-Gradient (PG) agent; proposed a novel Opponent-Modeling technique in MCTS that replaces random rollouts with supervised-learning inference to predict opponent moves from discard history.
High-Performance Computing (HPC), Core Member · 2024
- Achievement: lowest execution time among all teams — Champion of the 3rd NCHC Cup.
- Problem: optimizing molecular-dynamics simulations (NAMD) on the Taiwania 2 supercomputer.
- Methodology: analyzed bottlenecks for a 200-system amyloid-protein simulation; tuned CPU-affinity across 36 cores / 8 GPUs; parallelized to cut I/O overhead and maximize throughput for small-scale molecular systems.
Generative AI for Software-Engineering Education (Project GAIOOP) Lead Developer · Sep. 2023 – Aug. 2024
- Built a Web IDE integrating an LLM (ChatGPT), StarUML, and automated testing (JUnit) to generate scaffolded code via prompt engineering and fine-tuning.
- Ran A/B testing (independent-samples t-test), validating significant learning-outcome gains for lower-proficiency students.
Technical Skills
- Languages — C/C++, Python, Java, JavaScript, TypeScript
- Deep Learning & AI — PyTorch, MCTS/AlphaZero implementation, LLM fine-tuning, RAG
- HPC & Parallel Computing — MPI, OpenMP, CUDA, Slurm, performance tuning (CPU/GPU affinity)
- Cloud & DevOps — AWS Certified AI Practitioner, Docker, Kubernetes (K8s), Git, CI/CD
Teaching, other competitions, projects, sports, clubs, and personal challenges live on the Beyond page.