J48-Simplecart-Reptree: An Improved Ensemble ML Approach for Heart Disease Prediction

Cardiovascular diseases are a major concern of global health. A huge chunk of death rates all over the world is attributed to it. For this reason, early identification can be so important, as it can lead to required treatment and care measures and should minimize the impact of the disease. This research seeks to develop a heart disease classifier based on a dataset comprising 1024 patient histories with 14 clinical variables, including cholesterol levels, blood pressure, age, gender, ECG, and exercise-induced angina, amongst others. Three decision trees (DT) in the form of algorithms were employed, namely J48, SimpleCart, and REPTree, and yielded an accuracy of 98.05 %, <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{9 7. 0 7 \%}$</tex>, and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{8 4. 4 1 \%}$</tex> respectively. To increase the efficiency of the model, a new model that would combine those three algorithms has been created, using the majority vote method. The ensemble model is found to be better than the individual models with an accuracy of 99.02 %, a precision of 100 %, a recall of 97.98 %, and an F1 score of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{9 8. 9 8 \%}$</tex>. This research therefore establishes that ensemble learning models are business efficient when used in clinical diagnosis. The high level of accuracy combined with a nearly equal value of the entropy of measures used in the model indicates that it may be useful in practical applications for diagnostics in medicine to help the doctor make a diagnosis. Future works will concern the expansion of the patient record data set variety as well as the incorporation of real-time data to increase reliability and usefulness to clinical practice. In this respect, this paper provides the basis for the creation of automated diagnostic systems to tackle the global impact of heart disease.