Configurable Heuristic Adaptation for Improving Best First Search in AI Planning

Ivan Serina, Mauro Vallati

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Automated planning is one of the most prominent AI challenges. In the last few decades, there has been a great deal of activity in designing planning techniques and planning engines, with a focus on forward state-space search. Despite the ubiquitous use of heuristics in AI planning, these techniques are susceptible to being easily trapped by undetected dead ends and huge search plateaus. In this paper we introduce a highly configurable heuristic adaptation process based on the idea of dynamically penalising unpromising actions when an inconsistency in the heuristic evaluation is detected; its aim is to reduce the bias affecting specific actions, thereby encouraging exploration by the search process and adding diversity in the neighbourhood selection process. Our extensive experimental analysis demonstrates that the proposed heuristic can be configured to improve significantly the performance of best first search planning on a range of benchmark domains.
Original languageEnglish
Title of host publicationThe 32th International Conference on Tools with Artificial Intelligence (ICTAI) 2020
PublisherIEEE
Publication statusAccepted/In press - 2 Sep 2020
Event32nd International Conference on Tools with Artificial Intelligence - Online due to COVID-19
Duration: 9 Nov 202011 Nov 2020
Conference number: 32
https://ictai2020.org/index.html

Conference

Conference32nd International Conference on Tools with Artificial Intelligence
Abbreviated titleICTAI 2020
Period9/11/2011/11/20
Internet address

Fingerprint Dive into the research topics of 'Configurable Heuristic Adaptation for Improving Best First Search in AI Planning'. Together they form a unique fingerprint.

Cite this