1887

n South African Computer Journal - An expert system algorithm for the identifcation problem

USD

 

Abstract

An algorithm is proposed for an expert system to solve the general identifcation problem which is formally defined. the algorithm follows a statistical approach and is based on the work of the author in [2] and Vogts in [11]. A measure is defined to express the similarity between an observed specimen to be identified and any one of the possible classes. It is based on a similarity function of the common property values. This approach solves the inaccuracies of the linearized approach given in [2]. Using a measure of clustering, a method to select the 'best' next property to be established during the execution of the algorithm is proposed. A prototype of the algorithm is tested ona small knowledge base to identify South African cycads, similar to that given in [2]. the comparison with the results in [2]shows that the proposed algorithm is a significant improvement.

Loading

Article metrics loading...

/content/comp/2000/26/EJC27905
2000-11-01
2016-12-04
This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error