Abstract

Impaired heart function can have fatal consequences for human health, some heart defects can even cause death. This research aims to predict heart defects using the Backpropagation Neural Network machine learning algorithm approach. Apart from building artificial neural network models for predictive capabilities, data exploration is also carried out to gain insight and correlation for each variable or column in the dataset. Evaluate the model created using RMSE, confusion matrix, and ROV curve to find out how well the model performs in making predictions. The research results showed that the RMSE (Root Mean Square Error) value was 0.33, the confusion matrix value was 88.89%, and the ROC curve value with the ROC curve area = 0.88.