Align-RUDDER: Learning From Few Demonstrations by Reward Redistribution
Vihang Patil Affiliation: ELLIS Unit Linz and LIT AI Lab, Institute for Machine Learning, Johannes Kepler University Linz Correspondence to: Markus Hofmarcher Affiliation: ELLIS Unit Linz and LIT AI Lab, Institute for Machine Learning, Johannes Kepler University Linz Marius-Constantin Dinu Affiliation: ELLIS Unit Linz and LIT AI Lab, Institute for Machine Learning, Johannes Kepler University Linz Affiliation: Dynatrace Research Matthias Dorfer Affiliation: enliteAI Patrick Blies Affiliation: enliteAI Johannes Brandstetter Affiliation: ELLIS Unit Linz and LIT AI Lab, Institute for Machine Learning, Johannes Kepler University Linz Affiliation: Now at Microsoft Research José Arjona-Medina Affiliation: ELLIS Unit Linz and LIT AI Lab, Institute for Machine Learning, Johannes Kepler University Linz Affiliation: Dynatrace Research Sepp Hochreiter Affiliation: ELLIS Unit Linz and LIT AI Lab, Institute for Machine Learning, Johannes Kepler University Linz Affiliation: Institute of Advanced Research in Artificial Intelligence
Abstract
Reinforcement learning algorithms require many samples when solving complex hierarchical tasks with sparse and delayed rewards. For such complex tasks, the recently proposed RUDDER uses reward redistribution to leverage steps in the $Q$ -function that are associated with accomplishing sub-tasks. However, often only few episodes with high rewards are available as demonstrations since current exploration strategies cannot discover them in reasonable time. In this work, we introduce Align-RUDDER, which utilizes a profile model for reward redistribution that is obtained from multiple sequence alignment of demonstrations. Consequently, Align-RUDDER employs reward redistribution effectively and, thereby, drastically improves learning on few demonstrations. Align-RUDDER outperforms competitors on complex artificial tasks with delayed rewards and few demonstrations. On the Minecraft ObtainDiamond task, Align-RUDDER is able to mine a diamond, though not frequently. Code is available at github.com/ml-jku/align-rudder.
原文 arXiv:2009.14108;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2009.14108v2