Risk-Sensitive Reinforcement Learning, Mathematical Finance, and Machine Learning
My research lies at the intersection of risk measurement, decision-making under uncertainty, and machine learning. I focus on developing mathematically grounded approaches to risk-sensitive reinforcement learning, optimal decision-making, portfolio optimization, and credit risk assessment These efforts aim to advance both theoretical understanding and practical methods in finance, operations, and related applications.
Recent work has explored risk-sensitive reinforcement learning. In particular, we have developed distributional reinforcement learning methods for optimizing broad classes of risk measures, with an emphasis on theoretical guarantees, interpretable risk preferences, and reliable policy learning. This work connects mathematical risk theory with modern reinforcement learning and examines how risk preferences can be incorporated systematically into sequential decision-making.
My earlier work spans portfolio optimization and risk-sensitive financial decision-making, as well as machine-learning approaches to credit risk, including ensemble methods for corporate credit rating assessment and sequence-based clustering for long-term credit risk analysis.
