Abstract:

In order to improve the predicted accuracy of the classification based on genetic algorithm,the fitness functions for evaluating the rule equMity are discussed in this paper.A fitness function based on a weighted sum of confidence and suppo~ is then proposed,which can substitutes the traditional fitness function based on sensitivity and specificity.Both the theoretically an alytical an d the experimental results show that the propo sed fitness function has a greater advantage in the evolutionary search of predictive association rules over the traditional one,and is helpful to the improvement of the machine learning based on genetic algorithm.

Key words: machine leaning, evolutionary learning, genetic algorithm, association rule, classification, prediction