Mining useful Macro-actions in Planning

TL;DR

This paper introduces a data mining-based algorithm to identify useful macro-actions in planning, leading to significant improvements in classical benchmarks by enhancing plan synthesis efficiency.

Contribution

It presents a novel online macro-action learning algorithm that integrates data mining techniques into planning search processes.

Findings

Significant performance improvements on six classical planning benchmarks.

Effective identification of useful macro-actions through data mining.

Enhanced plan synthesis efficiency in planning algorithms.

Abstract

Planning has achieved significant progress in recent years. Among the various approaches to scale up plan synthesis, the use of macro-actions has been widely explored. As a first stage towards the development of a solution to learn on-line macro-actions, we propose an algorithm to identify useful macro-actions based on data mining techniques. The integration in the planning search of these learned macro-actions shows significant improvements over six classical planning benchmarks.