In this work, we study the use of the Bellman equation as a surrogate objective for value
prediction accuracy. While the Bellman equation is uniquely solved by the true value …

Sample efficiency and risk-awareness are central to the development of practical
reinforcement learning (RL) for complex decision-making. The former can be addressed by …

This note summarizes the optimization formulations used in the study of Markov decision
processes. We consider both the discounted and undiscounted processes under the …

In this paper, we address the problem of continuous-time reinforcement learning in
scenarios where the dynamics follow a stochastic differential equation. When the underlying …

This thesis investigates the intersection of reinforcement learning (RL) with function
approximation and limited data to develop practical, broadly applicable algorithms. Our …

A remarkable feature of an intelligent agent is the ability to make sequences of smart
decisions that are executed in coordination to reach goals. As can be seen by watching …

As the continuous limit of the gradient-based optimization algorithms, Fokker Planck (FP)
equation can provide a qualitative description of the algorithm's behavior and give principled …

In this work, we analyze the effectiveness of the Bellman equation as a proxy objective for
value prediction accuracy in off-policy evaluation. While the Bellman equation is uniquely …