Fuzzy rule extraction using hybrid evolutionary models for data mining systems

Data mining is a very popular technique which is successfully used in many areas. The aim of this paper is to present a Hybrid model for data classification from input datasets. The proposed model extracts knowledge using fuzzy rule based systems and performs classification task by fuzzy if-then rules. The proposed method performs the classification task and extracts required knowledge using fuzzy rule based systems which consists of fuzzy if-then rules. In order to do so the hybrid ant colony and simulated annealing algorithms have been used to optimize extracted fuzzy rule set. “ACSA”, a self development data mining software system based on swarm intelligence, is applied to experiment on eight data sets taken from UCI Repository on Machine Learning. The results illuminate the algorithm proposed in this paper has better performance in classification accuracy. The results are compared with those of well-known methods, and show the systems competitive efficiency.