Heart disease detection based on machine learning algorithms

Cardiovascular diseases are the first cause of death all over the world. By using artificial intelligence algorithms and, in particular, machine learning approaches it is possible to predict risky situations due to heart disease. Various approaches are investigated in this thesis such as Neural Network, Support Vector Machine, Decision Tree, Naive Bayes, Logistic Regression and Stochastic Gradient Descent to extract predictive models in order to test for the presence or absence of heart disease. Thanks to the public dataset from UCI, it is possible to take advantage of medical data to train the proposed models. A comparison among the different approaches based on the performance was included in this thesis. The tests of the proposed models revealed performances in terms of accuracy in the range 77\%-90.6\%. The Naive Bayes model has been the model with the highest accuracy (90.6\%), highest precision (96.4\%) and shortest time for classification (0.003 seconds).