A Review of Feature Selection and Classification Approaches for Heart Disease Prediction

Abstract

Cardiovascular disease has been the number one illness to cause death in the world for years. As information technology develops, many researchers have conducted studies on a computer-assisted diagnosis for heart disease. Predicting heart disease using a computer-assisted system can reduce time and costs. Feature selection can be used to choose the most relevant variables for heart disease. It includes filter, wrapper, embedded, and hybrid. The filter method excels in computation speed. The wrapper and embedded methods consider feature dependencies and interact with classifiers. The hybrid method takes advantage of several methods. Classification is a data mining technique to predict heart disease. It includes traditional machine learning, ensemble learning, hybrid, and deep learning. Traditional machine learning uses a specific algorithm. The ensemble learning combines the predictions of multiple classifiers to improve the performance of a single classifier. The hybrid approach combines some techniques and takes advantage of each method. Deep learning does not require a predetermined feature engineering. This research provides an overview of feature selection and classification methods for the prediction of heart disease in the last ten years. Thus, it can be used as a reference in choosing a method for heart disease prediction for future research.