AN APPLICATION OF NOVEL DISTINGUISHABILITY BASED WEIGHTED FEATURE SELECTION ALGORITHMS FOR IMPROVED CLASSIFICATION OF GENE MICROARRAY DATASET

Abstract

The experience of the past must justifiably and intelligently adopted for the present so that the future enjoys the enhanced happiness and comfort by mitigating the suffering due to disease as one part. Data mining applied over biometric knowledge could be effectively utilized to systematically diagnose a disease namely the colon tumor cancer and thereby scientifically treat with better understanding and accuracy. This research is one such an attempt to harness the experience of the past and the presence for the expectations of the future. newlineThe success of a machine learning algorithms depends on quality of the data. Hence,data mining play a vital role to explore, comprehend, analyze, understand, interpret and extract previously unknown, yet valid, useful and valuable information along with the knowledge discovery from microarray technology expression data. This include a search for genes that had similar or correlated patterns of expression. For that, the feature selection is one of the frequently used technique for data preprocessing. newlineEvery one irrespective of the position in the society is using vast data. It might be in the form of documents, or numerical formats or images. As the data available in the different formats, proper analyis is required not only to analyze these data but also to arrive at a good decision for giving treatment for the patient in mitigating the suffering even if it not to save their life. For that, many feature selection algorithms had been developed in order to extract the informative genes data especially in colon tumor data. newline newline