By Topic

A learning process of the matching identification problem

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$31 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

4 Author(s)
Kline, D. ; Dept. of Adm. Sci., Kent State Univ., OH, USA ; Dull, F. ; Mehrez, A. ; Steinberg, G.

Recently, Mehrez and Steinberg (1995) described and studied the matching identification problem (MIP). The MIP is a form of knowledge acquisition problem from the field of artificial intelligence. For instance, an expert system infers knowledge from a set of examples. But how do you most quickly acquire the examples that knowledge is inferred from? The MIP is a special case of this problem. Although an optimal algorithm was not found by Mehrez and Steinberg, they described two general types of heuristics. We describe in this paper an optimal algorithm for the case of K=2, and an improved heuristic for general K, which identifies a chosen subset with 6% fewer inquiries on average when N=15, K=3. The heuristic improves relative to the Type I heuristic as N increases, K held constant. The improved heuristic is concerned with the symbols yet unclassified as being in the chosen subset or not in the chosen subset. By inquiring subsets with all unclassified symbols, we most quickly “span” the set of unclassified numbers. Closed form equations are developed for the expected number of inquiries required and the variance of the number of inquiries required for the optimal algorithm. Computational studies are provided for Mehrez and Steinberg's Type I heuristics, the K=2 optimal algorithm, and the spanning heuristic

Published in:

Systems, Man, and Cybernetics, Part B: Cybernetics, IEEE Transactions on  (Volume:27 ,  Issue: 2 )