How To Learn How To Predict Heart Disease.
Yayıncı:
5. Uluslararası Muhendislik Mimarlık ve Tasarım Kongresi
Heart diseases occur with a problem that occurs in the vessels that go to the heart, openness in the heart cover or any unexpected place of the heart.There are two major factors that increase the risk of heart disease.These are genetic factors and environmental factors.The most common appearance in heart diseases is a heart attack due to plaque in the vessels.The most important symptom of heart disease is chest pain, and research shows that the average age of people who experience these pain decreases from day to day.According to the data of the World Health Organization, under the name of circulatory system diseases, 17.9 million people in the world per year die from heart disease, and in Turkey, according to the data of the statistical institution, an average of 168 thousand people die.Heart disease is a disease that has effects such as chest pain, breathing difficulty, swelling, tiredness and discomfort and will negatively affect your daily life.When diagnosed early, the disease does not pass to advanced stages and the start of treatment saves the patient's life.In this study, a set of data of 303 patients with heart disease, which consisted of 165 patients with heart disease, has been applied various machine learning methods with characteristics such as gender, diabetes, age, cholesterol, type of breast pain.Logistics regression, k-en close neighbor, support vector machines, naive bayes, decision tree, random forest, lightgbm model, xgboost model, ridge model and bagging model algorithms were compared, output results were evaluated and the accuracy value of 90.16% was obtained with random forest algorithm using different parameters.Keywords: prediction of heart disease, machine learning, random forest algorithm