Projects

Deterministically Constrained Stochastic Optimization
The problems appear widely in constrained M-estimation, semiparametric models, and constrained neural networks.
Deterministically Constrained Stochastic Optimization
Statistical Inference of Stochastic Approximation
We study the stationarity of SA methods, perform online hypothesis testing, and build online confidence intervals.
Statistical Inference of Stochastic Approximation
Long-horizon Nonlinear Optimal Control
We study the sensitivity of optimal control policy to the system perturbation, and propose offline distributed methods and online real-time MPC methods.
Long-horizon Nonlinear Optimal Control
Semiparametric and Graphical Models
We study the estimation of various semiparametric and latent-variable graphical models in high dimensions, with applications to genetics and neuroscience.
Semiparametric and Graphical Models