Selection of Optimal Set of Diagnostic Tests with Use of Evolutionary Approach in Intelligent Systems

This paper concerns problem of selection of optimal subset of irredundant unconditional diagnostic tests by means of evolutionary approach. The method of correction of features’ weight, cost and damage coefficients for the test patterns recognition is proposed. The suggestion is made that evolutionary programming approach would be more appropriate than genetic algorithm because of disadvantage of crossover use for multi-objective problems solution. Future research tasks to compare different evolutionary algorithms for solving problem at hand are outlined.

• Type: Conference
• Year: 2007
• Where: EUSFLAT
• Authors: A. E. Yankovskaya, Y. R. Tsoy
• Keywords: EUSFLAT 2007, Evolutionary Approach, Evolutionary Programming Approach, Paper Concerns Problem