A Performance Evaluation of Chi-Square Pruning Techniques in Class Association Rules Optimization

Abstract

Associative classification is recognized by its high accuracy and strong flexibility in managing unstructured data. However, the performance is still induced by low quality dataset which comprises of noised and distorted data during data collection. The noisy data affected support value of an itemset and so it influenced the performance of an associative classification. The performance of associative classification is relied on the classification where the classification is worked based on the class association rules which generated from frequent rule mining process. To optimize the frequent itemsets based on the support value, in this research, we proposed a new optimization pruning technique to prune decision tree according to the correlation of each decision tree branches using genetic algorithm.