Simultaneous clustering and feature ranking by competitive repetition suppressionlearning with application to gene data analysis
Davide Bacciu, Alessio Micheli, Antonina Starita · 2007
Abstract: The paper presents feature-wise Competitive Repetition-suppression (CoRe) clustering, a novel unsu-pervised algorithm that deals with the automatic deter-mination of the unknown cluster number and simultane-ous feature ranking. The proposed model addresses the limitations of the original CoRe learning algorithm when dealing with high dimensional data, extending the rep-etition suppression competition on a feature-wise basis. The effectiveness of the approach is tested on gene ex-pression data from DNA microarrays: the results show that the feature-wise CoRe clustering algorithm is able to detect the known data partitioning in a completely unsu-pervised fashion. Moreover, it simultaneously develops a gene ranking that is consistent with the state-of-the-art list of gene relevance for the selected benchmark datasets.