Fast TRAC: A Parameter-Free Optimizer for Lifelong Reinforcement Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Muppidi, Aneesh, Zhang, Zhiyu, Yang, Heng
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917822886576128
author Muppidi, Aneesh
Zhang, Zhiyu
Yang, Heng
author_facet Muppidi, Aneesh
Zhang, Zhiyu
Yang, Heng
contents A key challenge in lifelong reinforcement learning (RL) is the loss of plasticity, where previous learning progress hinders an agent's adaptation to new tasks. While regularization and resetting can help, they require precise hyperparameter selection at the outset and environment-dependent adjustments. Building on the principled theory of online convex optimization, we present a parameter-free optimizer for lifelong RL, called TRAC, which requires no tuning or prior knowledge about the distribution shifts. Extensive experiments on Procgen, Atari, and Gym Control environments show that TRAC works surprisingly well-mitigating loss of plasticity and rapidly adapting to challenging distribution shifts-despite the underlying optimization problem being nonconvex and nonstationary.
format Preprint
id arxiv_https___arxiv_org_abs_2405_16642
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Fast TRAC: A Parameter-Free Optimizer for Lifelong Reinforcement Learning
Muppidi, Aneesh
Zhang, Zhiyu
Yang, Heng
Machine Learning
Artificial Intelligence
A key challenge in lifelong reinforcement learning (RL) is the loss of plasticity, where previous learning progress hinders an agent's adaptation to new tasks. While regularization and resetting can help, they require precise hyperparameter selection at the outset and environment-dependent adjustments. Building on the principled theory of online convex optimization, we present a parameter-free optimizer for lifelong RL, called TRAC, which requires no tuning or prior knowledge about the distribution shifts. Extensive experiments on Procgen, Atari, and Gym Control environments show that TRAC works surprisingly well-mitigating loss of plasticity and rapidly adapting to challenging distribution shifts-despite the underlying optimization problem being nonconvex and nonstationary.
title Fast TRAC: A Parameter-Free Optimizer for Lifelong Reinforcement Learning
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2405.16642