Saved in:
Bibliographic Details
Main Authors: Ovinnikov, Ivan, Bykovets, Eugene, Buhmann, Joachim M.
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2409.08012
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917774732820480
author Ovinnikov, Ivan
Bykovets, Eugene
Buhmann, Joachim M.
author_facet Ovinnikov, Ivan
Bykovets, Eugene
Buhmann, Joachim M.
contents Inverse reinforcement learning methods aim to retrieve the reward function of a Markov decision process based on a dataset of expert demonstrations. The commonplace scarcity and heterogeneous sources of such demonstrations can lead to the absorption of spurious correlations in the data by the learned reward function. Consequently, this adaptation often exhibits behavioural overfitting to the expert data set when a policy is trained on the obtained reward function under distribution shift of the environment dynamics. In this work, we explore a novel regularization approach for inverse reinforcement learning methods based on the causal invariance principle with the goal of improved reward function generalization. By applying this regularization to both exact and approximate formulations of the learning task, we demonstrate superior policy performance when trained using the recovered reward functions in a transfer setting
format Preprint
id arxiv_https___arxiv_org_abs_2409_08012
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Learning Causally Invariant Reward Functions from Diverse Demonstrations
Ovinnikov, Ivan
Bykovets, Eugene
Buhmann, Joachim M.
Machine Learning
Artificial Intelligence
Inverse reinforcement learning methods aim to retrieve the reward function of a Markov decision process based on a dataset of expert demonstrations. The commonplace scarcity and heterogeneous sources of such demonstrations can lead to the absorption of spurious correlations in the data by the learned reward function. Consequently, this adaptation often exhibits behavioural overfitting to the expert data set when a policy is trained on the obtained reward function under distribution shift of the environment dynamics. In this work, we explore a novel regularization approach for inverse reinforcement learning methods based on the causal invariance principle with the goal of improved reward function generalization. By applying this regularization to both exact and approximate formulations of the learning task, we demonstrate superior policy performance when trained using the recovered reward functions in a transfer setting
title Learning Causally Invariant Reward Functions from Diverse Demonstrations
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2409.08012