English

Robust and Adaptive Temporal-Difference Learning Using An Ensemble of Gaussian Processes

Machine Learning 2021-12-03 v1 Machine Learning

Abstract

Value function approximation is a crucial module for policy evaluation in reinforcement learning when the state space is large or continuous. The present paper takes a generative perspective on policy evaluation via temporal-difference (TD) learning, where a Gaussian process (GP) prior is presumed on the sought value function, and instantaneous rewards are probabilistically generated based on value function evaluations at two consecutive states. Capitalizing on a random feature-based approximant of the GP prior, an online scalable (OS) approach, termed {OS-GPTD}, is developed to estimate the value function for a given policy by observing a sequence of state-reward pairs. To benchmark the performance of OS-GPTD even in an adversarial setting, where the modeling assumptions are violated, complementary worst-case analyses are performed by upper-bounding the cumulative Bellman error as well as the long-term reward prediction error, relative to their counterparts from a fixed value function estimator with the entire state-reward trajectory in hindsight. Moreover, to alleviate the limited expressiveness associated with a single fixed kernel, a weighted ensemble (E) of GP priors is employed to yield an alternative scheme, termed OS-EGPTD, that can jointly infer the value function, and select interactively the EGP kernel on-the-fly. Finally, performances of the novel OS-(E)GPTD schemes are evaluated on two benchmark problems.

Keywords

Cite

@article{arxiv.2112.00882,
  title  = {Robust and Adaptive Temporal-Difference Learning Using An Ensemble of Gaussian Processes},
  author = {Qin Lu and Georgios B. Giannakis},
  journal= {arXiv preprint arXiv:2112.00882},
  year   = {2021}
}
R2 v1 2026-06-24T08:00:40.728Z