Reinforcement Learning for Individual Optimal Policy from Heterogeneous Data

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Miao, Rui, Shahbaba, Babak, Qu, Annie
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917323545247744
author Miao, Rui
Shahbaba, Babak
Qu, Annie
author_facet Miao, Rui
Shahbaba, Babak
Qu, Annie
contents Offline reinforcement learning (RL) aims to find optimal policies in dynamic environments in order to maximize the expected total rewards by leveraging pre-collected data. Learning from heterogeneous data is one of the fundamental challenges in offline RL. Traditional methods focus on learning an optimal policy for all individuals with pre-collected data from a single episode or homogeneous batch episodes, and thus, may result in a suboptimal policy for a heterogeneous population. In this paper, we propose an individualized offline policy optimization framework for heterogeneous time-stationary Markov decision processes (MDPs). The proposed heterogeneous model with individual latent variables enables us to efficiently estimate the individual Q-functions, and our Penalized Pessimistic Personalized Policy Learning (P4L) algorithm guarantees a fast rate on the average regret under a weak partial coverage assumption on behavior policies. In addition, our simulation studies and a real data application demonstrate the superior numerical performance of the proposed method compared with existing methods.
format Preprint
id arxiv_https___arxiv_org_abs_2505_09496
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Reinforcement Learning for Individual Optimal Policy from Heterogeneous Data
Miao, Rui
Shahbaba, Babak
Qu, Annie
Machine Learning
Offline reinforcement learning (RL) aims to find optimal policies in dynamic environments in order to maximize the expected total rewards by leveraging pre-collected data. Learning from heterogeneous data is one of the fundamental challenges in offline RL. Traditional methods focus on learning an optimal policy for all individuals with pre-collected data from a single episode or homogeneous batch episodes, and thus, may result in a suboptimal policy for a heterogeneous population. In this paper, we propose an individualized offline policy optimization framework for heterogeneous time-stationary Markov decision processes (MDPs). The proposed heterogeneous model with individual latent variables enables us to efficiently estimate the individual Q-functions, and our Penalized Pessimistic Personalized Policy Learning (P4L) algorithm guarantees a fast rate on the average regret under a weak partial coverage assumption on behavior policies. In addition, our simulation studies and a real data application demonstrate the superior numerical performance of the proposed method compared with existing methods.
title Reinforcement Learning for Individual Optimal Policy from Heterogeneous Data
topic Machine Learning
url https://arxiv.org/abs/2505.09496