DocumentCode :
1331273
Title :
SHAPES: A novel approach for learning search heuristics in under-constrained optimization problems
Author :
Doan, Khanh P V ; Wong, Kit Po
Author_Institution :
Dept. of Electr. & Electron. Eng., Western Australia Univ., Nedlands, WA, Australia
Volume :
9
Issue :
5
fYear :
1997
Firstpage :
731
Lastpage :
746
Abstract :
Although much research in machine learning has been carried out on acquiring knowledge for problem-solving in many problem domains, little effort has been focused on learning search-control knowledge for solving optimization problems. This paper reports on the development of SHAPES, a system that learns heuristic search guidance for solving optimization problems in intractable, under-constrained domains based on the Explanation-Based Learning (EBL) framework. The system embodies two new and novel approaches to machine learning. First, it makes use of explanations of varying levels of approximation as a mean for verifying heuristic-based decisions, allowing heuristic estimates to be revised and corrected during problem-solving. The provision of such a revision mechanism is particularly important when working in intractable and under-constrained domains, where heuristics tend to be highly over-generalized, and hence at times will give rise to incorrect results. Second, it employs a new linear and quadratic programming-based weight-assignment algorithm formulated to direct search toward optimal solutions under best-first search. The algorithm offers a direct method for assigning rule strengths and, in so doing, avoids the need to address the credit-assignment problem faced by other iterative weight-adjustment methods
Keywords :
explanation; knowledge acquisition; knowledge based systems; learning (artificial intelligence); optimisation; problem solving; quadratic programming; SHAPES; best-first search; explanation-based learning framework; explanations; heuristic search guidance; knowledge acquisition; machine learning; problem-solving; quadratic programming-based weight-assignment algorithm; search heuristics learning; search-control knowledge; under-constrained optimization problems; Cost function; Iterative algorithms; Iterative methods; Machine learning; Machine learning algorithms; Problem-solving; Production systems; Shape;
fLanguage :
English
Journal_Title :
Knowledge and Data Engineering, IEEE Transactions on
Publisher :
ieee
ISSN :
1041-4347
Type :
jour
DOI :
10.1109/69.634752
Filename :
634752
Link To Document :
بازگشت