Comparative analogy on classification and clustering of genomic signal by a novel factor analysis and f-score method

Abstract

Data is the centroid in any dimension of scientific research. Most of the research arena is problem specific with respect to its data. In machine learning more and more data samples are used for training a machine for efficient cluster formation, prediction, recognition, classification. Variation between data sample to data sample and data sample from its centroid using Euclidean distance, Mahanalabolis distance for better cluster formation has better focused among computational researchers. In this paper, we proposed a novel loss less feature extraction and selection using feedforward cascaded Factor Analysis and F-score method. Comparative analysis of classification and clustering after efficient feature extraction is applied to life killer cancer disease. Tumor classification is based on multilayer perceptron and clustering is based on k-means clustering method.