Improving the Performance of Machine Learning Algorithms for Heart Disease Diagnosis by Optimizing Data and Features

Heart is one of the most important members of the body, and heart disease is the major cause of death in the world and Iran. This is why the early/on time diagnosis is one of the significant basics for preventing and reducing deaths of this disease. So far, many studies have been done on heart disease with the aim of prediction, diagnosis, and treatment. However, most of them have been mostly focused on the prediction of heart disease. The purpose of this study is to develop models for heart disease diagnosis using machine learning, neural network, and deep learning algorithms. The models have been developed using the Cleveland heart disease dataset from University of California Irvine (UCI) repository. After complete data processing, including outlier detection, normalization, discretization, feature selection and feature extraction, the dataset is transformed into two normalized data and discretized data, according to the nature of the algorithms. Moreover, in constructing models of machine learning and neural networks, two randomized searches with cross-validation and grid search with Talos scan approaches are used for model tuning. Among evaluated models, including decision tree algorithms, random forest, support vector machine (SVM) and XGBoost, the highest accuracy is 92.9% using SVM, and among neural network models, multilayer perceptron (MLP) has resulted in the highest accuracy of 94.6%.