Computational Intelligence Techniques for Classification in Microarray Analysis

Abstract

During the last few years there has been a growing need for using computational intelligence techniques to analyze microarray data. The aim of the system presented in this study is to provide innovative decision support techniques for classifying data from microarrays and for extracting knowledge about the classification process. The computational intelligence techniques used in this chapter follow the case-based reasoning paradigm to emulate the steps followed in expression analysis. This work presents a novel filtering technique based on statistical methods, a new clustering technique that uses ESOINN (Enhanced Self-Organizing Incremental Neuronal Network), and a knowledge extraction technique based on the RIPPER algorithm. The system presented within this chapter has been applied to classify CLL patients and extract knowledge about the classification process. The results obtained permit us to conclude that the system provides a notable reduction of the dimensionality of the data obtained from microarrays. Moreover, the classification process takes the detection of relevant and irrelevant probes into account, which is fundamental for subsequent classification and an extraction of knowledge tool with a graphical interface to explain the classification process, and has been much appreciated by the human experts. Finally, the philosophy of the CBR systems facilitates the resolution of new problems using past experiences, which is very appropriate regarding the classification of leukemia.