A Genetic algorithm aided hyper parameter optimization based ensemble model for respiratory disease prediction with Explainable AI

• DOI: 10.1371/journal.pone.0308015
• Publication Date: 2024-12-02T18:39:34Z
• Authors: Balraj Preet Kaur; Harpreet Singh; Rahul Hans; Sanjeev Kumar Sharma; Chetna Sharma; Md. Mehedi Hassan
• Keywords: Hyperparameter; Hyperparameter Optimization; AdaBoost; Feature (linguistics); Ensemble Learning

In the current era, a lot of research is being done in domain disease diagnosis using machine learning. recent times, one deadliest respiratory diseases, COVID-19, which causes serious damage to lungs has claimed lives globally. Machine learning-based systems can assist clinicians early disease, reduce deadly effects disease. For successful deployment these systems, hyperparameter-based optimization and feature selection are important issues. Motivated by above, this proposal, we design an improved model predict existence among patients incorporating hyperparameter selection. To optimize parameters learning algorithms, with genetic algorithm proposed size set, performed binary grey wolf algorithm. Moreover, enhance efficacy predictions made hyperparameter-optimized models, ensemble stacking classifier. Also, explainable AI was incorporated define importance making use Shapely adaptive explanations (SHAP) values. experimentation, publicly accessible Mexico clinical dataset COVID-19 used. The results obtained show that superior prediction accuracy comparison its counterparts. all adaboost outperformed other algorithms. various performance assessment metrics, including accuracy, precision, recall, AUC, F1-score, were used assess results.