Optimizing ECG Heartbeat Classification with Improved Genetic Algorithm and Stacking Ensembles

The analysis of electrocardiogram (ECG) data presents substantial challenges as a result of the complexity of the data and the presence of noise, despite its widespread use in the diagnosis of heart conditions. In this investigation, we suggest a novel methodology that integrates a Stacking ensemble learning framework with an enhanced Genetic Algorithm (GA) for hyperparameter optimization to improve the automatic classification of ECG signals. Although traditional GAs are effective, they frequently fail to achieve global optimum solutions due to limited population diversity, particularly when relying heavily on mutation. To overcome this constraint, we implement improvements to the GA’s selection, crossover, mutation, and termination processes, guaranteeing optimal outcomes and efficient resource utilization. Experimental evaluations indicate that the proposed Stacking ensemble method, which employs Logistic Regression as the meta-model, achieves an F1 score of 93.60% and an accuracy of 98.85%. These results emphasize the potential of combining enhanced GAs with Stacking ensemble learning to achieve robust and accurate ECG signal classification, thereby providing a promising solution for automated cardiac diagnostics. The source code and implementation details are publicly available at: https://github.com/Thangnezzz/Improved-Genetic-Algorithm-and-Stacking-Ensembles .