Predicting healthcare utilizing new machine learning techniques and metaheuristic algorithms

Predicting disease outcomes remains challenging. This research integrates machine learning with metaheuristic optimization to improve diagnostic accuracy. The authors present a predictive system that merges RFC and GBC with bio-inspired optimization methods, ARO, and DOA. The optimization methods serve to establish optimal hyperparameter values. These lead to improved model performance. Experimental outcomes on a synthetic healthcare database demonstrate that the RFAR model (RFC optimized with ARO) achieves the highest accuracy (0.989), followed by GBAR (GBC + ARO) and RFDO (RFC + DOA). Combining ensemble learning with metaheuristics improves predictive performance over standard methods, supporting automated diagnostic systems and personalized healthcare planning.