Comparative Analysis of Data Mining Algorithms for Cancer Gene Expression Data

Preeti Thareja,Rajender Singh Chhillar

doi:10.14569/ijacsa.2021.0121035

Preeti Thareja, Rajender Singh Chhillar

Open Access

https://doi.org/10.14569/ijacsa.2021.0121035

Copy DOI

Abstract

Cancer is amongst the most challenging disorders to diagnose nowadays, and experts are still struggling to detect it on early stage. Gene selection is significant for identifying cancer-causing different parameters. The two deadliest cancers namely, colorectal cancer and breast malignant, is found in male and female, respectively. This study aims at predicting the cancer at an early stage with the help of cancer bioinformatics. According to the complexity of illness metabolic rates, signaling, and interaction, cancer bioinformatics is among strategies to focus bioinformatics technologies like data mining in cancer detection. The goal of the proposed study is to make a comparison between support vector machine, random forest, decision tree, artificial neural network, and logistic regression for the prediction of cancer malignant gene expression data. For analyzing data against algorithms, WEKA is used. The findings show that smart computational data mining techniques could be used to detect cancer recurrence in patients. Finally, the strategies that yielded the best results were identified.

Highlights

Non - communicable diseases (NCDs) responsible for 71 percent of all fatalities worldwide
The findings show that the k-nearest neighbor (kNN) classifier outperforms the DT classifier in the Breast Cancer (BC) classification
This work is carried to provide a brief outline of the state of art techniques SVM, ANN, Logistic Regression, Decision Tree and Random Forest applied on datasets for classification

Summary

Introduction

Non - communicable diseases (NCDs) responsible for 71 percent of all fatalities worldwide. Non-communicable diseases (NCDs) are illnesses which are never spread by pathogens. They are long-term illnesses with a sluggish course that are caused by a mix of biological, physiologic, ecological, and behavioral variables. Malignant is a non-communicable disorder in wherein some tissues grow out of control and extend to many other areas of the organism. Malignant is another name used for cancer that can begin practically at any place in the trillions of cells that make up the human body. Data analysis has a remarkable ability to uncover hidden patterns in disease prediction [2,3]

Objectives

Methods

Results

Conclusion