Detection of Common Risk Factors Leading to the Cardiovascular Illness Using Machine Learning

Cases of heart illness are developing at an alarming rate, and it is critical and important to predict any such ailments in advance. This is a challenging task that must be completed accurately and effectively. The research paper focuses on the development of a machine learning-based artificial intelligence-based cardiac disease diagnosis system. We demonstrate how machine learning can assist in predicting whether a person will develop cardiovascular disease using medical records of patients. Moreover, the research highlights the most important leading causes of cardiovascular diseases. To predict and classify the patient with heart disease, we applied various machine learning methods such as Gaussian, KNN, XGBoost and Decision trees. To tackle the problem of unbalanced data, different techniques were applied, like Nearmiss, random undersampling, random oversampling, and ENN combined with random undersampling. The proposed model’s strength was quite gratifying, as it was able to predict evidence of having a heart illness in a specific individual by using XGBoost with ENN random undersampling, which demonstrated good accuracy in contrast to the other commonly used classifiers such as balancing data techniques. We evaluate the proposed model by using sratified cross validation and various metrics like accuracy, recall, f-score, precision, and ROC curve. Furthermore, we use the SHAP value to identify major risk factors for cardiovascular disease.