Early heart disease prediction using machine learning Algorithms

Heart disease ranks among the most perilous ailments of our time, taking a toll on thousands of lives a year. Detection in its early stages can be lifesaver. Machine learning is a subset of artificial intelligence that aids in the rapid, efficient, and cost-effective prediction of heart disease. In this research, several ML models were trained handling the Cleveland heart disease dataset to determine which performed well in predicting heart disease cases. The selection of data attributes and the choice of ML algorithm have a great impact on the performances of different algorithms. The four machine learning models tested were Random Forest, Decision Tree, Logistic Regression, and Naive Bayes. This work has shown the Random Forest classifier to be the best one as far as the accuracy is concerned at 89.2%, where it is closely followed by the Naïve Bayes at 87%. Both logistic regressions achieved reasonably high accuracy at 85.2%, respectively. The decision tree was the least accurate of the others, with an accuracy of 82% and overall applicability in this case. Their comparisons were done based on various evaluation metrics such as accuracy, precision, recall, F-measure, and error rate, where Random Forest outperformed the others. This study brings to light the capabilities of machine-learning techniques in the early detection of heart disease. Random Forest appears to be the best-performing model. This study shows that ML can indeed be a great help for medical practitioners in their diagnostic procedures by refining them to enhance early detection and subsequent outcomes for patients.