COSBO: Conservative Offline Simulation-Based Policy Optimization

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kargar, Eshagh, Kyrki, Ville
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914954639048704
author Kargar, Eshagh
Kyrki, Ville
author_facet Kargar, Eshagh
Kyrki, Ville
contents Offline reinforcement learning allows training reinforcement learning models on data from live deployments. However, it is limited to choosing the best combination of behaviors present in the training data. In contrast, simulation environments attempting to replicate the live environment can be used instead of the live data, yet this approach is limited by the simulation-to-reality gap, resulting in a bias. In an attempt to get the best of both worlds, we propose a method that combines an imperfect simulation environment with data from the target environment, to train an offline reinforcement learning policy. Our experiments demonstrate that the proposed method outperforms state-of-the-art approaches CQL, MOPO, and COMBO, especially in scenarios with diverse and challenging dynamics, and demonstrates robust behavior across a variety of experimental conditions. The results highlight that using simulator-generated data can effectively enhance offline policy learning despite the sim-to-real gap, when direct interaction with the real-world is not possible.
format Preprint
id arxiv_https___arxiv_org_abs_2409_14412
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle COSBO: Conservative Offline Simulation-Based Policy Optimization
Kargar, Eshagh
Kyrki, Ville
Machine Learning
Artificial Intelligence
Robotics
Offline reinforcement learning allows training reinforcement learning models on data from live deployments. However, it is limited to choosing the best combination of behaviors present in the training data. In contrast, simulation environments attempting to replicate the live environment can be used instead of the live data, yet this approach is limited by the simulation-to-reality gap, resulting in a bias. In an attempt to get the best of both worlds, we propose a method that combines an imperfect simulation environment with data from the target environment, to train an offline reinforcement learning policy. Our experiments demonstrate that the proposed method outperforms state-of-the-art approaches CQL, MOPO, and COMBO, especially in scenarios with diverse and challenging dynamics, and demonstrates robust behavior across a variety of experimental conditions. The results highlight that using simulator-generated data can effectively enhance offline policy learning despite the sim-to-real gap, when direct interaction with the real-world is not possible.
title COSBO: Conservative Offline Simulation-Based Policy Optimization
topic Machine Learning
Artificial Intelligence
Robotics
url https://arxiv.org/abs/2409.14412