Heart disease prediction with machine learning algorithms

Nowadays, with the development of much faster computers and machine algorithm systems, the use of artificial intelligence in the healthcare sector has increased significantly. The use of these algorithms in the field of medicine has led to significant developments. Artificial intelligence techniques are frequently used in the diagnosis and treatment of heart diseases. In the thesis study, using machine learning, one of the sub-branches of artificial intelligence, a data set of heart disease was taken as a classification problem and a prediction was made on the probability of the patients in this data set to have heart disease. The obtained data set was trained with the sub-algorithms of the supervised learning system, a separate model was created with each classification algorithm, and then heart disease was predicted by comparing these created models with the test data set. KNIME Analytics Platform was used as the application. Among the classification algorithms, LPROP Multilayer Algorithm, K-Means Clustering, Support Vector Machines, Naive Bayes, Decision Tree, Logistic Regression, neural network algorithms and collective learning methods were used. When the accuracy rate of the test data set was evaluated according to the results obtained in the trials and tests carried out with the application, the highest accuracy rate was achieved with the Navi Bayes method with a rate of 85.26%. In the study, the error matrices of the predictions were also evaluated and shown in tables