Skip to content
conftrace_
2023 AAAI AAAI 2023

A New Challenge in Policy Evaluation

Abstract

Abstract This paper proposes a new challenge in policy evaluation: to improve the online data efficiency of Monte Carlo methods via information extracted from offline data while maintaining the unbiasedness of Monte Carlo methods.

🌉 Interdisciplinary Bridge - Artificial Intelligence and Reinforcement Learning

Authors

Shangtong Zhang

Topics

Machine Learning > Optimization & Theory > Statistical Learning Reinforcement Learning > Methods > Policy Learning

Keywords

policy evaluation unbiased estimation monte carlo method offline data online data efficiency
Download PDF

More from AAAI 2023

All AAAI 2023 papers →
The Analysis of Deep Neural Networks by Information Theory: From Explainability to Generalization
Towards Societal Impact of AI
Information Transfer in Multitask Learning, Data Augmentation, and Beyond
Building Compositional Robot Autonomy with Modularity and Abstraction
Accurate Detection of Weld Seams for Laser Welding in Real-World Manufacturing
Blending Advertising with Organic Content in E-commerce via Virtual Bids