I am an NSF TRIPODS Postdoctoral Research Fellow at University of Illinois at Urbana-Champaign, working with Prof. R. Srikant and Prof. Niao He. My research interests include reinforcement learning, neural networks and stochastic control.

Recently, I have been working on the theoretical analysis of reinforcement learning algorithms with neural network approximation. Check out my Google Scholar profile for a list of my recent papers.
News
  • "Linear Convergence of Entropy-Regularized Natural Policy Gradient with Linear Function Approximation", joint work with Niao He and R. Srikant, is now available. [link]

  • "Sample Complexity and Overparameterization Bounds for Temporal Difference Learning with Neural Network Approximation", joint work with Siddhartha Satpathi, Niao He and R. Srikant, is now available. [link]

  • "A Lyapunov-Based Methodology for Constrained Optimization with Bandit Feedback", joint work with Yilin Zheng and Atilla Eryilmaz, is now available. [link]

  • "Group-Fair Online Allocation in Continuous Time", joint work with Swati Gupta and Atilla Eryilmaz, was published in the Proceedings of NeurIPS 2020. [link]

  • Does restarting a prolonged task expedite the completion and improve time-efficiency? How can you learn optimal restart strategies with bandit feedback? "Continuous-Time Multi-Armed Bandits with Controlled Restarts", joint work with Atilla Eryilmaz and R. Srikant, is now available. [link]

  • "Budget-Constrained Bandits over General Cost and Reward Distributions", joint work with Atilla Eryilmaz and R. Srikant, was published in the Proceedings of AISTATS 2020. [link][slides]