My current research interests can be broadly classified as "physics-informed artificial intelligence." They include:
Learning for convex and combinatorial optimization
Distributional robustness and rare events
Risk-sensitive stochastic optimal control and reinforcement learning
Reduced order models for natural gas dynamics
Bayesian approaches to generative artificial intelligence
I work on a variety of networked systems that arise in engineering, primarily power transmission grids and natural gas networks. Recently, I have become interested in power distribution grids and microgrids.