Heart disease prediction using machine learning models

Abstract

As heart disease is getting more and more attention from people and with the development of machine learning. The author can use machine learning techniques to construct models to analyze patients' physical and psychological conditions and make prediction about if or not a patient has heart disease or not based on the patient's personal information it takes. The dataset is a collection of patients' personal information that have different genders, race, physical condition and psychological condition and so on. Our experiment starts with data preprocessing by trimming off extreme values and incomplete values while balancing the number of data presented by different classes. The author uses four models to predict whether a patient is experiencing heart disease or not which can help patients' to know their status and alert people not to get heart disease. The author starts with KNN, ANN then support vector machine and eventually Random Forest, the author tune model's hyperparameters to make sure they are at their best state. At last the author will compare the accuracy with each other to find out which model is efficient and accurate in predicting heart disease.