Combining Macro-operators with Control Knowledge

Abstract

Inductive Logic Programming (ilp) methods have proven to succesfully acquire knowledge with very different learning paradigms, such as supervised and unsupervised learning or relational reinforcement learning. However, very little has been done on applying it to General Problem Solving (gps). One of the ilp-based approaches applied to gps is hamlet. This method learns control rules (heuristics) for a non linear planner, prodigy4.0, which is integrated into the ipss system; control rules are used as an effective guide when building the planning search tree. Other learning approaches applied to planning generate macro-operators, building high-level blocks of actions, but increasing the branching factor of the search tree. In this paper, we focus on integrating the two different learning approaches (hamlet and macro-operators learning), to improve a planning process. The goal is to learn control rules that decide when to use the macro-operators. This process is successfully applied in several classical planning domains.