Pessimistic Nonlinear Least-Squares Value Iteration for Offline Reinforcement Learning

Authors: Qiwei Di, Heyang Zhao, Jiafan He, Quanquan Gu

ICLR 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical In this section, we prove an instance-dependent regret bound of Algorithm 1. Our algorithmic design comprises three innovative components: (1) a variance-based weighted regression scheme that can be applied to a wide range of function classes, (2) a subroutine for variance estimation, and (3) a planning phase that utilizes a pessimistic value iteration approach. Our algorithm enjoys a regret bound that has a tight dependency on the function class complexity and achieves minimax optimal instance-dependent regret when specialized to linear function approximation.
Researcher Affiliation Academia Qiwei Di1, Heyang Zhao1, Jiafan he1, Quanquan Gu1 1Department of Computer Science, University of California, Los Angeles {qiwei2000,hyzhao,jiafanhe19,qgu}@cs.ucla.edu
Pseudocode Yes Algorithm 1 Pessimistic Nonlinear Least-Squares Value Iteration (PNLSVI)
Open Source Code No The paper does not provide any explicit statements about releasing source code or links to a code repository.
Open Datasets No The paper describes using a 'batch-dataset D' for offline RL, but does not name any specific public datasets or provide access information for any dataset used.
Dataset Splits No The paper is theoretical and does not report on experiments with dataset splits, so no training/validation/test splits are mentioned.
Hardware Specification No The paper is theoretical and does not report on experiments, thus no hardware specifications are provided.
Software Dependencies No The paper is theoretical and does not report on experiments, thus no software dependencies with version numbers are listed.
Experiment Setup No The paper focuses on theoretical algorithm design and analysis, and does not provide details about an experimental setup, such as hyperparameters or training settings.