Postdoc, University of Maryland, College Park
2 papers at NeurIPS 2025
A robust RL framework for quantitative trading with introducing novel elliptic uncertainty sets to model the market impact.
We propose a class of unbiased zeroth-order gradient estimator.