Partially Observable Reinforcement Learning with Memory Traces

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Eberhard, Onno, Muehlebach, Michael, Vernade, Claire
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912410536771584
author Eberhard, Onno
Muehlebach, Michael
Vernade, Claire
author_facet Eberhard, Onno
Muehlebach, Michael
Vernade, Claire
contents Partially observable environments present a considerable computational challenge in reinforcement learning due to the need to consider long histories. Learning with a finite window of observations quickly becomes intractable as the window length grows. In this work, we introduce memory traces. Inspired by eligibility traces, these are compact representations of the history of observations in the form of exponential moving averages. We prove sample complexity bounds for the problem of offline on-policy evaluation that quantify the return errors achieved with memory traces for the class of Lipschitz continuous value estimates. We establish a close connection to the window approach, and demonstrate that, in certain environments, learning with memory traces is significantly more sample efficient. Finally, we underline the effectiveness of memory traces empirically in online reinforcement learning experiments for both value prediction and control.
format Preprint
id arxiv_https___arxiv_org_abs_2503_15200
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Partially Observable Reinforcement Learning with Memory Traces
Eberhard, Onno
Muehlebach, Michael
Vernade, Claire
Machine Learning
Partially observable environments present a considerable computational challenge in reinforcement learning due to the need to consider long histories. Learning with a finite window of observations quickly becomes intractable as the window length grows. In this work, we introduce memory traces. Inspired by eligibility traces, these are compact representations of the history of observations in the form of exponential moving averages. We prove sample complexity bounds for the problem of offline on-policy evaluation that quantify the return errors achieved with memory traces for the class of Lipschitz continuous value estimates. We establish a close connection to the window approach, and demonstrate that, in certain environments, learning with memory traces is significantly more sample efficient. Finally, we underline the effectiveness of memory traces empirically in online reinforcement learning experiments for both value prediction and control.
title Partially Observable Reinforcement Learning with Memory Traces
topic Machine Learning
url https://arxiv.org/abs/2503.15200