WOA-COVID-19: Whale Optimization Algorithm for Selection of Multi-Examination Features based on COVID-19 Infections

Background: As of late, COVID-19 has emerged as the most significant illness from 2019 onward. Millions of individuals have passed away, and many more are in dire circumstances. The different COVID-19 assessment aspects led to a variety of sophisticated diagnosis approaches. So, the true challenge is to figure out how to determine the best examination aspects that facilitate the diagnosis process. Objective: This study proposes a Whale Optimization Algorithm (WOA) for selection of multi-examination features based on COVID-19 infections. Methods: A dataset of 78 patients with 25 examination features of COVID-19 (demographic, laboratory findings, vital signs, symptoms, and chronic conditions) are used in the feature selection process. The feature selection process is based on adapting the single objective technique of WOA. K-Nearest Neighbors (KNN) algorithm is used for classification of COVID-19 patients into three main cases which are mild, moderate and severe. Six feature selection metaheuristic algorithms have used during evaluation process which are Particle Swarm Optimization (PSO), Multi-Verse Optimizer (MVO), Grey Wolf Optimizer (GWO), Moth-Flame Optimization (MFO), Firefly Algorithm(FFA),and BAT algorithm). Results: In our research, we found that the WOA algorithm yields a smaller set of selected features—just six out of 25 features are chosen as the foundation for the classification phase. Furthermore, the classification accuracy reaches 92.5 %. The robustness of WOA-Covid-19 is high as seen by the tiny standard deviations compared to other methods when applied to the specified dataset. Conclusion: The final findings show that the suggested approach enhances the efficiency of WOA-Covid-19 in addressing feature selection issues.