Research Interests

Machine Learning: Provable methods for deep learning and representation learning, subspace learning, multiview learning, streaming algorithms for kernel methods, online learning

Stochastic Optimization: Non-convex optimization, stochastic approximation for large-scale problems, robust adversarial learning

Differential Privacy: Computational tradeoffs in private machine learning, local learning, federated learning, privacy in streaming algorithms and continual release models

My research is supported by an NSF CAREER award on Understanding Inductive Biases in Modern Machine Learning, an NSF BIGDATA award on Privacy in Machine Learning, a DARPA award on Robust Adversarial Learning, an NSF BIGDATA award on Stochastic Approximation for Subspace and Multiview Representation Learning, an NSF TRIPODS award on Foundations of Graph and Deep Learning, and an NSF CRCNS award on Computational Neuroscience. See here for more details.