New Approaches for Meta-Heuristic Frameworks: A Position Paper

David A. Ostrowski, George Schleis

This paper summarizes our position on the use of AI-based techniques applied to the development and support of metaknowledge. Our intent is to derive higher level abstractions using current heuristic methods to the goal of plan development and selection. A generalized framework is discussed, allowing for development of strategic plans coupled with an interchangeable learning mechanism. Along with this framework we present potential issues and conditions to address during the development phases that are of importance to the institutionalization of these concepts. We feel that the highest potential exists at the determination of meta-level (control and strategic) knowledge — specifically using a hybrid approach incorporating current expert system paradigms in conjunction with Machine Learning (EC, NN) techniques.

Subjects: 3. Automated Reasoning; 12. Machine Learning and Discovery

Submitted: May 9, 2008

This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.