A Planner with Quality Goal and Its Speed-up Learning for Optimization Problem

Aimsof traditional planners had beenlimited to finding a sequenceof operators rather than finding an optimal or neax-optimalfinal state. Consequent]y, the performanceimprovementsystems combinedwith the planners had only aimed at efficiently finding an axbitraxysolution, but not necessarily the optimal solution. However,there axe manydomainswherewecall for quality of the final state for each problem. In this paper, we proposean extension of a planner for optimization problems, and another application of EBL to problemsolving: learning control knowledge to improveseaxching performancefor an optimal or a near-optimal solution by analyzing reasons a solution is better than another. Theproposed method was applied to technology mapping in LSI design, a typical optimization problem. The system with the learned control knowledgesynthesizes optimal circuits four times faster than that without the control knowledge.

• Type: Conference
• Year: 1994
• Where: AIPS
• Authors: Masahiko Iwamoto
• Keywords: AIPS 1994, Artificial Intelligence, Control Knowledge, Optimal Solution, Optimization Problem