A Fuzzy Clustering Algorithm Based on Artificial Immune Principles

The Fuzzy C-Means algorithm (FCM) is a widely applied clustering method. However, it is usually trapped into the local optimum. In addition, its performance is very sensi- tive to the initialization. This paper proposes a new fuzzy clustering method based on the immune clonal selection principle, namely CFCM. The clonal selection algorithm is first used to optimize the number of fuzzy cluster cen- ters. The FCM is next employed for clustering the input data. Simulation results demonstrate that our novel ap- proach can overcome the drawbacks of the regular FCM with an improved data clustering performance.