Optimizing Cardiovascular Risk Assessment with a Soft Voting Classifier Ensemble0

According to the latest data from the World Health Organization (WHO), heart disease has been the leading cause of death worldwide for the past several decades. It includes a variety of conditions that affect the heart. In Pakistan heart disease claims the lives of at least thirty people every hour. The best-known application of artificial intelligence is machine learning (ML). It is linked to numerous heart disease risk factors and the necessity of time to acquire sensitive accurate and dependable methods in order to make an early diagnosis. Experimental options have included the UCI repository’s datasets on heart disease (which have 14 attributes) and cardiovascular diseases (12 attributes). The proposed ensemble soft voting classifier employs an ensemble of seven machine learning algorithms to provide binary classification, the Naïve Bayes K Nearest Neighbor SVM Kernel Decision Tree Random Forest Logistic Regression and Support Vector Classifier. The accuracy precision recall and F1_score value is provided by the suggested ensemble method with 70.9% 72.3% 68.6%, 70.1% and Random Forest gives 71.5%, 72.2%, 70.3%, and 71.2%. Rest of classifiers gave average scores. It means the proposed method provided best results while compared with Decision tree, Logistic regression, Support Vector Classifier (SVC), SVM Kernel, K Nearest Neighbor and Naïve Bayes. Only Random forest gives more accuracy than proposed method on cardio heart disease dataset.