A Comparative Analysis of Various Machine Learning Techniques in Diagnosis of Heart Disease

Nowadays, heart illness is somewhat common truth. Both the death rate and frequency are rising daily. In this study, three models-a simple logistic model using linear regression, K-Nearest Neighbors (K-NN), and Support Vector Machine (SVM)-are used for training and testing data in order to accurately predict heart disease. Following preprocessing, a ratio was used to divide the data into train and test sets. The results showed that the Support Vector Machine (SVM) approach was the most accurate model in terms of prediction accuracy, while the K-Nearest Neighbors (K-NN) and Logistic model with linear regression were relatively lesser accurate with respect to the same value of all parameters. To put it succinctly, these three prediction models were able to accurately forecast heart disease and exceeded accuracy. These findings suggest that these models are very practical and effective, and they can give physicians valuable information to help them identify and treat patients with heart disease more accurately. Dhaka Univ. J. Sci. 73(2): 171-175, 2025 (July)