Apurv Shukla
Apurv is a Postdoctoral Associate at University of Michigan working with Prof. Vijay Subramanian on dynamic multi-agent learning. His research focuses on designing algorithmic frameworks for decision-making in dynamic environments when learning agents are constrained by information asymmetry, preference heterogeneity, and privacy. Previously, he was at Texas A&M working with Prof. Le Xie and Prof. P.R. Kumar. He earned his PhD from Columbia University in 2022.
His representative publications are:
- Preference-based Pure Exploration preprint
NeurIPS 2024 with Debabrota Basu - Differentially Private Online Resource Allocation with P.R. Kumar
- Ignorance is bliss: Stochastic lookahead economic dispatch under forecast errors, with Le Xie and P.R. Kumar
- Sample Complexity of Chance Constrained Optimization in Dynamic Environment preprint
ACC 2024 with Qian Zhang and Le Xie - Robust Streaming PCA preprint
NeurIPS 2022 with Daniel Bienstock, Minchan Jeong and Seyoung Yun
Teaching
- ECEN 713 Data Sciences and Applications for Modern Power Systems, Fall 2023