Integrating planning and learning: the PRODIGY architecture

TLDR

Planning is a complex reasoning task suited for improving performance and knowledge through learning by accumulating and interpreting experience, and PRODIGY integrates planning with multiple learning mechanisms. The paper aims to describe the PRODIGY planner, detail learning modules and two methods for generating higher‑quality plans, illustrate them with examples, present preliminary results, and discuss the architecture’s evolution toward a robust integrated planning and learning system. Learning in PRODIGY occurs at planner decision points, using mutually interpretable knowledge structures, and the architecture incorporates several learning modules and two methods for generating higher‑quality plans. Preliminary empirical results demonstrate the effectiveness of the introduced techniques for generating higher‑quality plans.

Abstract

Abstract Planning is a complex reasoning task that is well suited for the study of improving performance and knowledge by learning, i.e. by accumulation and interpretation of planning experience. PRODIGY is an architecture that integrates planning with multiple learning mechanisms. Learning occurs at the planner's decision points and integration in PRODIGY is achieved via mutually interpretable knowledge structures. This article describes the PRODIGY planner, briefly reports on several learning modules developed earlier along the project, and presents in more detail two recently explored methods to learn to generate plans of better quality. We introduce the techniques, illustrate them with comprehensive examples, and show preliminary empirical results. The article also includes a retrospective discussion of the characteristics of the overall PRODIGY architecture and discusses their evolution within the goal of the project of building a large and robust integrated planning and learning system.