A Multi-objective Cluster Algorithm Based on GEP

Clustering is one of the main methods in data mining. Many clustering algorithms have been proposed so far. Among them, GEP-Cluster, a single-objective clustering algorithm, can automatically cluster with unknown clustering number. However, it is difficult for GEP-Cluster to find the high-quality solution in the limited search space. Aiming at the problems, a multi-objective clustering algorithm based on gene expression programming, MOGEP-Cluster, is proposed in this paper. To validate the effectiveness of MOGEP-Cluster, a set of experiments are performed on 5 benchmark datasets. The experimental results show that MOGEP-Cluster can find better solutions than GEP-Cluster.