MHER: Model-based Hindsight Experience Replay
arXiv:2107.00306
Abstract
Solving multi-goal reinforcement learning (RL) problems with sparse rewards is generally challenging. Existing approaches have utilized goal relabeling on collected experiences to alleviate issues raised from sparse rewards. However, these methods are still limited in efficiency and cannot make full use of experiences. In this paper, we propose Model-based Hindsight Experience Replay (MHER), which exploits experiences more efficiently by leveraging environmental dynamics to generate virtual achieved goals. Replacing original goals with virtual goals generated from interaction with a trained dynamics model leads to a novel relabeling method, model-based relabeling (MBR). Based on MBR, MHER performs both reinforcement learning and supervised learning for efficient policy improvement. Theoretically, we also prove the supervised part in MHER, i.e., goal-conditioned supervised learning with MBR data, optimizes a lower bound on the multi-goal RL objective. Experimental results in several point-based tasks and simulated robotics environments show that MHER achieves significantly higher sample efficiency than previous model-free and model-based multi-goal methods.
arXiv commentNeurIPS 2021 Workshop DeepRL
References in corpus (8)
- Continuous control with deep reinforcement learning
- Learning to reinforcement learn
- Maximum Entropy-Regularized Multi-Goal Reinforcement Learning
- Maximum Entropy Gain Exploration for Long Horizon Multi-goal Reinforcement Learning
- Generalized Hindsight for Reinforcement Learning
- Policy Continuation with Hindsight Inverse Dynamics
- Addressing Sample Complexity in Visual Tasks Using HER and Hallucinatory GANs
- MapGo: Model-Assisted Policy Optimization for Goal-Oriented Tasks