Hierarchical Meta-Reinforcement Learning via Automated Macro-Action Discovery

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Cho, Minjae, Sun, Chuangchuang
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866915066153009152
author Cho, Minjae
Sun, Chuangchuang
author_facet Cho, Minjae
Sun, Chuangchuang
contents Meta-Reinforcement Learning (Meta-RL) enables fast adaptation to new testing tasks. Despite recent advancements, it is still challenging to learn performant policies across multiple complex and high-dimensional tasks. To address this, we propose a novel architecture with three hierarchical levels for 1) learning task representations, 2) discovering task-agnostic macro-actions in an automated manner, and 3) learning primitive actions. The macro-action can guide the low-level primitive policy learning to more efficiently transition to goal states. This can address the issue that the policy may forget previously learned behavior while learning new, conflicting tasks. Moreover, the task-agnostic nature of the macro-actions is enabled by removing task-specific components from the state space. Hence, this makes them amenable to re-composition across different tasks and leads to promising fast adaptation to new tasks. Also, the prospective instability from the tri-level hierarchies is effectively mitigated by our innovative, independently tailored training schemes. Experiments in the MetaWorld framework demonstrate the improved sample efficiency and success rate of our approach compared to previous state-of-the-art methods.
format Preprint
id arxiv_https___arxiv_org_abs_2412_11930
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Hierarchical Meta-Reinforcement Learning via Automated Macro-Action Discovery
Cho, Minjae
Sun, Chuangchuang
Machine Learning
Artificial Intelligence
Meta-Reinforcement Learning (Meta-RL) enables fast adaptation to new testing tasks. Despite recent advancements, it is still challenging to learn performant policies across multiple complex and high-dimensional tasks. To address this, we propose a novel architecture with three hierarchical levels for 1) learning task representations, 2) discovering task-agnostic macro-actions in an automated manner, and 3) learning primitive actions. The macro-action can guide the low-level primitive policy learning to more efficiently transition to goal states. This can address the issue that the policy may forget previously learned behavior while learning new, conflicting tasks. Moreover, the task-agnostic nature of the macro-actions is enabled by removing task-specific components from the state space. Hence, this makes them amenable to re-composition across different tasks and leads to promising fast adaptation to new tasks. Also, the prospective instability from the tri-level hierarchies is effectively mitigated by our innovative, independently tailored training schemes. Experiments in the MetaWorld framework demonstrate the improved sample efficiency and success rate of our approach compared to previous state-of-the-art methods.
title Hierarchical Meta-Reinforcement Learning via Automated Macro-Action Discovery
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2412.11930