Saved in:
Bibliographic Details
Main Authors: Banerjee, Debangshu, Gopalan, Aditya
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2310.09358
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917595506016256
author Banerjee, Debangshu
Gopalan, Aditya
author_facet Banerjee, Debangshu
Gopalan, Aditya
contents Parametric, feature-based reward models are employed by a variety of algorithms in decision-making settings such as bandits and Markov decision processes (MDPs). The typical assumption under which the algorithms are analysed is realizability, i.e., that the true values of actions are perfectly explained by some parametric model in the class. We are, however, interested in the situation where the true values are (significantly) misspecified with respect to the model class. For parameterized bandits, contextual bandits and MDPs, we identify structural conditions, depending on the problem instance and model class, under which basic algorithms such as $ε$-greedy, LinUCB and fitted Q-learning provably learn optimal policies under even highly misspecified models. This is in contrast to existing worst-case results for, say misspecified bandits, which show regret bounds that scale linearly with time, and shows that there can be a nontrivially large set of bandit instances that are robust to misspecification.
format Preprint
id arxiv_https___arxiv_org_abs_2310_09358
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Bad Values but Good Behavior: Learning Highly Misspecified Bandits and MDPs
Banerjee, Debangshu
Gopalan, Aditya
Machine Learning
Artificial Intelligence
Parametric, feature-based reward models are employed by a variety of algorithms in decision-making settings such as bandits and Markov decision processes (MDPs). The typical assumption under which the algorithms are analysed is realizability, i.e., that the true values of actions are perfectly explained by some parametric model in the class. We are, however, interested in the situation where the true values are (significantly) misspecified with respect to the model class. For parameterized bandits, contextual bandits and MDPs, we identify structural conditions, depending on the problem instance and model class, under which basic algorithms such as $ε$-greedy, LinUCB and fitted Q-learning provably learn optimal policies under even highly misspecified models. This is in contrast to existing worst-case results for, say misspecified bandits, which show regret bounds that scale linearly with time, and shows that there can be a nontrivially large set of bandit instances that are robust to misspecification.
title Bad Values but Good Behavior: Learning Highly Misspecified Bandits and MDPs
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2310.09358