Abstract
Heart disease remains a predominant cause of mortality globally and in Indonesia, impacting both older and younger demographics due to pervasive unhealthy lifestyles, heredity, and lack of awareness of heart disease. This study addresses the critical problem of necessitating early detection methods to mitigate severe complications and fatalities associated with heart disease. With that, it is crucial to develop a robust and highly accurate prediction model for heart disease by integrating Artificial Neural networks (ANN) with Genetic Algorithms (GA). The model starts by constructing an ANN model utilizing the Keras Framework for streamlined training, followed by hyperparameter optimization through GA. As a result, this research found that the integrated ANN and GA model attains superior predictive accuracy, with the optimal configuration achieving an accuracy of 85.33%, precision of 92.55%, recall of 81.32%, and f1-score of 86.57% via Latin Hypercube Sampling (LHS). These show that the combination of ANN and GA can significantly increase prediction accuracy and model efficiency, as a solution for more effective heart disease identification at an early stage.