Machine learning-based approach for diagnosis of heart diseases

Heart-related diseases have become the main cause of death in the world in recent years and have emerged as the most life-threatening disease not only in our country but also in the whole world. Therefore, a reliable, accurate and applicable system is needed to ensure timely diagnosis of such diseases for appropriate treatment. Artificial intelligence makes important predictions in the healthcare sector using big data analysis and advanced algorithms. In recent years, artificial intelligence has been widely used in the health sector and provides many advantages. In the thesis study, early diagnosis and detection of heart diseases in new patients were achieved by using machine learning algorithms with high accuracy rates. In this context, it was observed that high success rates were detected in predicting the heart disease of patients by applying machine learning techniques to the data set consisting of 526 heart patients and 499 healthy samples and 14 attributes such as age, gender, chest pain type, etc. Machine Learning Algorithms K-Nearest Neighbor (KNN), Logistic Regression, Decision Tree, Support Vector Machines, Naive Bayes, Random Forest algorithms were used and the accuracy, precision, sensitivity and F Criterion values of the algorithms were included. In the obtained results, 205 test data constituting 20% of the total data were evaluated and 98.54% accuracy value was obtained with the Random Forest Algorithm