A Study of Machine Learning-based Models for Detecting Heart Disease

Heart failure stands among the leading causes that lead to worldwide deaths, while significantly adding to global disease statistics. Multiple risk factors, consisting of bad diet habits combined with unhealthy lifestyle choices and inherited predispositions, elevate the chance of developing cardiovascular diseases. Emerging medical diagnostic technologies, AI, and ML have shown recent effectiveness in the field. This study explores using various methods using machine learning for predicting heart disease and improving early detection. Methods like XG Boost, Logistic Regression, Support Vector Machine, Decision Trees, etc., indicate the probability or risk of suffering heart disease. In the following experimental section, various algorithms have been tested and compared with each other. Heart failure has presently become a pretty dangerous medical condition due to this reason. Therefore, a cross-validation technique was employed in this work, which improved the prediction accuracy for each of the three models under comparison. Between support vector machines and SMOTE-XGBoost, the results indicated that the logistic regression classifier would be the most effective machine learning model for predicting heart disease. It reached a figure as high as 92.59% accuracy with the logistic regression.