Case-Based Planning: An Integrated Theory of Planning, Learning and Memory
Abstract
This dissertation presents a theory of case-based planning that integrates memory oriented learning with an active planner. Case-based planning requires a machine planner that makes use of its own past experience in developing new plans. A case-based planner relies on memory instead of a base of rules. Memories of past successes are accessed and modified to create new plans. Memories of past failures are used to warn the planner of impending problems, and memories of past repairs are called upon to tell the planner how to how to deal with them. This view of planning from experience is supported by a learning system that incorporates new experiences into the planner's episodic memory. This learning algorithm gains from the planner's failures as well as its successes. Successful plans are stored in memory, indexed by the goals they satisfy and the problems they avoid. Failures are also stored, indexed by the features in the world that predict them. By storing failures as well as successes, the planner is able to anticipate and avoid future plan failures. A process of plan repair is also presented in which plan failures are diagnosed through a causal analysis of the steps and states that led to their occurrence. This causal analysis is used to access repair strategies for the general situation. These strategies are then transformed into specific alterations for the faulty plan at hand. This theory improves on past planning models in three areas: failure avoidance, plan repair and plan reuse.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1986
- Accession Number
- ADA180271
Entities
People
- Kristian J. Hammond
Organizations
- Yale University