IMMUNE SYSTEM-BASED CLUSTERING AND CLASSIFICATION ALGORITHMS

• Published in: Mathematical Methods In Scattering Theory And Biomedical Engineering pp. 399 - 406
• Main Authors: SOTIROPOULOS, D. N., TSIHRINTZIS, G. A.
• Format: Book Chapter
• Language: English
• Published: WORLD SCIENTIFIC 01.08.2006
• Subjects: Biomedical Engineering
• ISBN: 9812568603; 9812773193; 9789812773197; 9789814477598; 9814477591; 9789812568601
• DOI: 10.1142/9789812773197_0039

Artificial Immune Systems (AIS) comprise adaptive systems based upon functions, principles and models of theoretical immunology, that can be applicable to a wide range of subjects, providing metaphors for the development of high level abstractions of functions or mechanisms that can be utilized for solving real world problems. Clearly, AlS-based clustering and classification algorithms are characterized by an ability to adapt their behavior so as to cope efficiently with extremely complex and continuously changing environments. In this paper, we develop an Artificial Immune Network (AIN) for clustering a set of unlabelled multidimensional feature vectors. Next, we assess the ability of the network to classify feature vectors that were not used in its generation phase. Finally, we utilize agglomerative algorithms based on graph theory and especially on the minimum spanning tree for the visualization of the AIN and the intrinsic clusters of the data set.