Importance Weighted Transfer of Samples in Reinforcement Learning

TL;DR

This paper introduces a model-based importance weighting method for transferring experience samples across tasks in reinforcement learning, improving learning efficiency and robustness against negative transfer.

Contribution

It proposes a novel importance weighting technique that automatically estimates sample relevance, enhancing transfer learning in RL with a finite-sample analysis and empirical validation.

Findings

Outperforms state-of-the-art transfer methods in RL.

Robust to negative transfer even with dissimilar source tasks.

Provides a finite-sample theoretical analysis of the approach.

Abstract

We consider the transfer of experience samples (i.e., tuples < s, a, s', r >) in reinforcement learning (RL), collected from a set of source tasks to improve the learning process in a given target task. Most of the related approaches focus on selecting the most relevant source samples for solving the target task, but then all the transferred samples are used without considering anymore the discrepancies between the task models. In this paper, we propose a model-based technique that automatically estimates the relevance (importance weight) of each source sample for solving the target task. In the proposed approach, all the samples are transferred and used by a batch RL algorithm to solve the target task, but their contribution to the learning process is proportional to their importance weight. By extending the results for importance weighting provided in supervised learning literature, we develop a finite-sample analysis of the proposed batch RL algorithm. Furthermore, we empirically compare the proposed algorithm to state-of-the-art approaches, showing that it achieves better learning performance and is very robust to negative transfer, even when some source tasks are significantly different from the target task.