International Research Journal of Education and Technology

-Cardiovascular disease remains one of the most pressing global health challenges, accounting for nearly 17.9 million deaths annually according to the World Health Organization. Early and accurate prediction of heart disease can substantially reduce mortality rates by enabling timely clinical intervention. This research investigates the application of eight machine learning classification algorithms — namely Logistic Regression, K-Nearest Neighbors, Support Vector Machine, Decision Tree, Random Forest, Gradient Boosting, Naive Bayes, and Artificial Neural Network — to predict the presence or absence of heart disease in patients using a combination of clinical and demographic risk factors.

The study employs the widely used UCI Cleveland Heart Disease dataset comprising 303 patient records with 14 attributes. A systematic pipeline encompassing data preprocessing, feature engineering, model training, hyperparameter tuning via Grid Search Cross-Validation, and performance evaluation using multiple metrics — accuracy, precision, recall, F1-score, and AUC-ROC — was implemented. Experimental results demonstrate that the Random Forest classifier achieved the highest overall accuracy of 91.8% and an AUC-ROC score of 0.956, outperforming all other algorithms. This paper provides a thorough comparative analysis alongside statistical significance testing to identify the most suitable algorithm for clinical decision support. The findings offer meaningful insights for healthcare practitioners and data scientists working toward intelligent diagnostic systems.

Key Words:Heart disease prediction, machine learning, classification algorithms, random forest, feature selection, clinical decision support, AUC-ROC, UCI dataset.