A Clustering Model Inspired by Humoral Immunity

In biological immune system, B-cells secrete large numbers of antibodies to recognize and eliminate the antigens. Inspired by the relationship of B-cells and antibodies, an effective immune model is presented in this paper. As its learning capability, this model can recognize not only the existing antigens but also the antigens that are unknown. The structure of the model and the detailed algorithm are given in this paper. And the validity of the model is proved through an experiment of motor fault data clustering. Keywords-artificial immune system; clustering; B-cell; antibody I. INTRODUCTION Currently, information technology develops very fast. So, huge information is produced, and data mining can transform them into useful knowledge. Clustering is an important domain of data mining. It can find out the distributing rule of data character through comparing the comparability and diversity of data, and help researchers to obtain more profound comprehension and cognition (1). But the traditional clustering algorithm are deficient on clustering precision and convergent speed, such as k-means algorithm, Bayesian learning algorithm, fuzzy C means algorithm (FCM), etc.