WiOpt 2021 Workshops
Workshop on Reinforcement Learning and Stochastic Control in Queues and Networks
Approximate planning and learning for partially observed systems
that this approach works well in practice. Joint work with Jayakumar Subramanian, Amit Sinha, and Raihan Seraj.
Online Reinforcement Learning for MDPs and POMDPs via Posterior Sampling
Learning algorithm for optimal network control
Data-Driven Stochastic Network Control via Reinforcement Learning
Given a stable policy, we further develop a model-based RL method and prove that it converges to the optimal policy. Our method demonstrates promising performance in a variety of network control problems including routing, dynamic server allocation and switch scheduling.
A Provably-Efficient Model-Free Algorithm for Constrained Markov Decision Processes
Learning based meta-scheduling in wireless networks
Job Dispatching Policies for Queueing Systems with Unknown Service Rates
Gauri Joshi and Weina Wang