Research
Publications
-
2025Optimizing DRAM Chip Treatment Selection by Enhanced Robust Reinforcement Learning with Decision Regime in Progressive Decision-Making Data. · IEEE IEEM 2025 · Oral Presentation · Melbourne, Australia
Working Papers
-
2026✍️ A Robust Trust-Aware Algorithm for Quantitative Strategies. Developing a robust, trustworthy quant trading algorithm with formal reliability guarantees — extending prior RL robustness work to quantitative finance. [Work in progress]
Honors & Awards
-
2025President & CEO Award, Samsung Electronics · All Departments
Selected as one of only three project teams company-wide among 100,000+ employees for the most outstanding technical achievement of the year. Recognized for establishing a core DRAM optimization technology and achieving record-high manufacturing yields through ML-driven automation. -
2024Best Paper Award, Samsung Electronics · Memory Business Internal Technical Conference
Honored for research excellence in "Optimizing DRAM Chip Selection via Robust RL". -
2023DREAM Fair Top 10, Samsung Electronics · Memory Business
Selected as Top 10 among 1,000+ participants across all new BS/MS/PhD graduate engineers. Recognized for building 7 automated DRAM analysis tools and a TB-scale data engineering framework. -
2023Best Seminar Award · Development Best Performer, Samsung Electronics · DRAM Product Engineering Team
Best Seminar Award for ML for Semiconductor Analysis instructor role (2022–24). Development Best Performer for 7 automation tools that reduced data preprocessing time by ~90%.
Research Interests
- Robust / Trustworthy MLRobustness under distribution shift; reliable decision-making under uncertainty.
- Reinforcement LearningRobust RL for progressive decision-making data; regime-aware policy switching.
- Causal InferenceSelective interaction-aware causal inference via DML for semiconductor process optimization.
- Quantitative FinanceAI-driven strategy research; HMM regime detection; trust-scored signal engines.
- ML for SemiconductorsHybrid ML for automated parameter tuning; yield optimization at mass-production scale.