Abstract

The study explores the utilization of machine learning model in predicting heart disease ,employing pearson’s chi
square test to assess feature importance and the Shapiro-Wilk test to ensure data normality .Eight machine
learning algorithm were evaluated including k nearest neighbour’s classifier, Logistic Regression, XGB Classifier,
Random Forest classifier, Gradient Boosting Classifier, Decision Tree Classifier, and SVC. The Gradient Boosting
Classifier achieved the highest accuracy (91.26%) while XGB Classfier showed the highest F-score
(0.1636).However recall score across most models found to be low ,highlighting challenges in identifying positive
cases effectively .By integrating statically test with machine learning the works gives a robust framework for the
early detection of heart disease .