Improvisation of Decision Tree Classification Performance in Breast Cancer Diagnosis using Elephant Herding Optimization

Abstract

A recent analysis found that ductal carcinoma, another name for breast cancer, is increasingly prevalent in women any time after puberty. Their brain, bones, liver, lungs, and other organs might acquire cancer as a result of their negligence during that particular time period. Hence to diagnose the breast cancer, the Decision Tree classifier can be implemented on the gene expression data. To enhance the results provided by decision tree classifier, the Elephant Herding Optimization will be used to transform the input gene expression data in this work. Principal Component Analysis is utilized for decreasing the dimensionality of gene expression data since the dimensionality of original dataset is very huge. The experiment is carried out on the dataset downloaded from the CuMiDa website. Through experiments it is found that, transform based on Elephant Herding Optimization helps the decision tree classifier for providing improved performance.