search query: @author Liang, T. P. / total: 3
reference: 1 / 3
« previous | next »
Author: | Liang, T. P. |
Title: | A composite approach to inducing knowledge for expert systems design. |
Journal: | Management Science
1992 : JAN, VOL. 38:1, p. 1-17 |
Index terms: | KNOWLEDGE-BASED SYSTEMS |
Language: | eng |
Abstract: | The process of developing an expert system (ES) includes acquiring knowledge from human experts, representing and organizing the knowledge, storing it in a knowledge base. Knowledge acquisition is a bottleneck for ES design. A new induction approach called composite rule induction system (CRIS) is presented. The benefit of this approach, the basic elements of CRIS, a series of empirical comparisons of the predictive validity of CRIS versus two entropy-based induction methods, statistical discrimination analysis and the propagation method in neural networks are discussed. It is shown that CRIS has higher predictive validity. |
« previous | next »
SCIMA