MAICS: Multilevel Artificial Immune Classification System
This paper presents a novel approach to feature selection and multiple-class classification problems. The proposed method is based on metaphors derived from artificial immune systems, clonal and negative selection paradigms. A novel clonal selection algorithm – Immune K-Means, is proposed. The proposed system is able to perform feature selection and model identification tasks by evolving specialized subpopulations of T- and B-lymphocytes. Multilevel evolution and real-valued coding enable for further extending of the proposed model and interpreting the subpopulations of lymphocytes as sets of evolving fuzzy rules.
KeywordsFeature Selection Fuzzy Rule Negative Selection Binary String Clonal Selection
Unable to display preview. Download preview PDF.
- 2.Bereta, M., Burczyński, T.: Hybrid immune algorithm for feature selection and classification of ECG signals. In: Burczyński, T., Cholewa, W., Moczulski, W. (eds.) Recent Developments in Artificial Intelligence Methods, Gliwice. AI-METH Series, pp. 25–28 (2005)Google Scholar
- 3.Goodman, D., Boggess, L., Watkins, A.: Artificial immune system classification of multiple-class problems. In: Artificial Neural Networks in Engineering (ANNIE 2002) (2002)Google Scholar