Teaching a robot tasks of arbitrary complexity via human feedback

Guan Wang, Carl Trimbach, Jun Ki Lee, Mark K. Ho, Michael L. Littman

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

7 Scopus citations

Abstract

This paper addresses the problem of training a robot to carry out temporal tasks of arbitrary complexity via evaluative human feedback that can be inaccurate. A key idea explored in our work is a kind of curriculum learning-training the robot to master simple tasks and then building up to more complex tasks. We show how a training procedure, using knowledge of the formal task representation, can decompose and train any task efficiently in the size of its representation.We further provide a set of experiments that support the claim that non-expert human trainers can decompose tasks in a way that is consistent with our theoretical results, with more than half of participants successfully training all of our experimental missions. We compared our algorithm with existing approaches and our experimental results suggest that our method outperforms alternatives, especially when feedback contains mistakes.

Original languageEnglish
Title of host publicationHRI 2020 - Proceedings of the 2020 ACM/IEEE International Conference on Human-Robot Interaction
Pages649-657
Number of pages9
ISBN (Electronic)9781450367462
DOIs
StatePublished - 9 Mar 2020
Event15th Annual ACM/IEEE International Conference on Human Robot Interaction, HRI 2020 - Cambridge, United Kingdom
Duration: 23 Mar 202026 Mar 2020

Publication series

NameACM/IEEE International Conference on Human-Robot Interaction
ISSN (Electronic)2167-2148

Conference

Conference15th Annual ACM/IEEE International Conference on Human Robot Interaction, HRI 2020
Country/TerritoryUnited Kingdom
CityCambridge
Period23/03/2026/03/20

Keywords

  • Human-robot interaction
  • Learning from human feedback
  • Linear temporal logic
  • Reinforcement learning

Fingerprint

Dive into the research topics of 'Teaching a robot tasks of arbitrary complexity via human feedback'. Together they form a unique fingerprint.

Cite this