Abstract
Heart disease is one of the main health problems worldwide, and heart disease prediction based on artificial neural networks (ANNs) has become a research hotspot, aiming to improve the accuracy of early diagnosis. At present, heart disease prediction still faces problems such as data complexity, feature selection, overfitting, etc., which need to be overcome in ANN models. The heart disease prediction based on ANN has potential importance, which can assist medical decision-making, reduce the incidence rate of heart disease, and provide support for personalized treatment and health management. This study proposes a heart disease diagnosis method relied on machine learning, which balances the dataset using the SMOTE (Synthetic Minority Over Sampling Technique) algorithm as well as ENN (Edited Nearest Neighbors) algorithm. Then, research was conducted based on a variety of machine learning models, such as LR (logistic regression), DT (decision tree), RF (random forest), GBDT (gradient enhancement decision tree), XGBoost (extreme gradient enhancement), SVM (support vector machine), and ANN (deep neural network), applied to predict heart disease datasets. Patients can measure the risk of heart disease through past medical history, household equipment, and personal habits. Research has shown that the auc and recall based on the ANN model are as high as 0.808 and 0.81, respectively.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have