Feature Selection Using Correlation Matrix on Metagenomic Data with Pearson Enhancing Inflammatory Bowel Disease Prediction

Abstract

The Fourth Industrial Revolution has brought up a vast amount of new innovative implementations. These can be used for numerous areas to make wealth and to improve human ways of living. Our point of view is to consider medical problems for enhancing prediction first. In this study, we like to rise a question about whether if we could enhancing Inflammatory Bowel Disease (IBD) prediction for early detect related sickness by feature selection on metagenomic data. Over the last few years, the prediction has been a challenge. Because of rare information and lacking data, the problem is not well considered enough. To bring back the subject, in this work, we propose a new way of enhancing Inflammatory Bowel Disease (IBD) prediction by using the Correlation Matrix with the Pearson on Metagenomic Data. Our implications have the purpose of finding out whether we could do predictions better using a specific amount of features selected by Pearson correlation coefficient. The result with the proposed method is quite promising, when we address some high correlation features out, the model can predict better comparing to randomly select features.