Efficient Deep Reinforcement Learning via Adaptive Policy Transfer

Abstract

Transfer Learning (TL) has shown great potential to accelerate Reinforcement Learning (RL) by leveraging prior knowledge from past learned policies of relevant tasks. Existing transfer approaches either explicitly computes the similarity between tasks or select appropriate source policies to provide guided explorations for the target task. However, how to directly optimize the target policy by alternatively utilizing knowledge from appropriate source policies without explicitly measuring the similarity is currently missing. In this paper, we propose a novel Policy Transfer Framework (PTF) to accelerate RL by taking advantage of this idea. Our framework learns when and which source policy is the best to reuse for the target policy and when to terminate it by modeling multi-policy transfer as the option learning problem. PTF can be easily combined with existing deep RL approaches. Experimental results show it significantly accelerates the learning process and surpasses state-of-the-art policy transfer methods in terms of learning efficiency and final performance in both discrete and continuous action spaces.

Accepted by IJCAI'2020

References in corpus (9)

• Continuous control with deep reinforcement learning
• Playing Atari with Deep Reinforcement Learning
• DeepMind Control Suite
• Policy Distillation
• Kickstarting Deep Reinforcement Learning
• Learnings Options End-to-End for Continuous Action Tasks
• Learning to Compose Skills
• The Termination Critic
• An Optimal Online Method of Selecting Source Policies for Reinforcement Learning