Adaptive Policy Regularization for Offline-to-Online Reinforcement Learning in HVAC Control
Published in NeurIPS CCAI & ACM BuildSys'24, 2024
Recommended citation: Liu, Hsin-Yu. Adaptive Policy Regularization for Offline-to-Online Reinforcement Learning in HVAC Control. NeurIPS CCAI & ACM BuildSys'24. Nov. 2024. https://dl.acm.org/doi/pdf/10.1145/3671127.3698163
This paper proposes an adaptive policy regularization approach for transferring policies from offline datasets to online deployment in HVAC control. The method uses weighted increased simple moving average Q-value estimators to stabilize policy updates and improve safety during online fine-tuning.
Results show significant improvements over baseline methods in HVAC control benchmarks.
title: "Adaptive Policy Regularization for Offline-to-Online Reinforcement Learning in HVAC Control"
collection: publications
permalink: /publication/adaptive-policy-regularization-2024
date: 2024-11-01
venue: "NeurIPS CCAI & ACM BuildSys'24"
paperurl: ''
paperurl: 'https://dl.acm.org/doi/pdf/10.1145/3671127.3698163'
citation: "Liu, Hsin-Yu. Adaptive Policy Regularization for Offline-to-Online Reinforcement Learning in HVAC Control. NeurIPS CCAI & ACM BuildSys'24. Nov. 2024."
This paper proposes an adaptive policy regularization approach for transferring policies from offline datasets to online deployment in HVAC control. The method uses weighted increased simple moving average Q-value estimators to stabilize policy updates and improve safety during online fine-tuning.
Results show significant improvements over baseline methods in HVAC control benchmarks.