Transfer Value Iteration Networks

TL;DR

This paper introduces Transfer VINs (TVINs), a transfer learning method that enables value iteration networks to generalize policies across domains with different action and feature spaces, improving performance with limited data.

Contribution

The paper proposes Transfer VINs (TVINs), a novel transfer learning approach that enhances VINs' ability to generalize across domains with domain-specific actions and features.

Findings

TVINs outperform VINs in cross-domain tasks

Performance gains are consistent across environments and hyperparameters

TVINs require limited training data for effective transfer

Abstract

Value iteration networks (VINs) have been demonstrated to have a good generalization ability for reinforcement learning tasks across similar domains. However, based on our experiments, a policy learned by VINs still fail to generalize well on the domain whose action space and feature space are not identical to those in the domain where it is trained. In this paper, we propose a transfer learning approach on top of VINs, termed Transfer VINs (TVINs), such that a learned policy from a source domain can be generalized to a target domain with only limited training data, even if the source domain and the target domain have domain-specific actions and features. We empirically verify that our proposed TVINs outperform VINs when the source and the target domains have similar but not identical action and feature spaces. Furthermore, we show that the performance improvement is consistent across different environments, maze sizes, dataset sizes as well as different values of hyperparameters such as number of iteration and kernel size.