Dynamic Scale Factor Based Elephant Herding Optimization with Weighted Support Vector Machine for COVID-19 Prediction

The rapid spread of Corona virus disease 2019 (COVID-19) makes its detection among the greatest challenges worldwide. A recent report indicates that COVID-19 has been detected in over 1.6 million individuals across several countries. To predict the risk of COVID-19, predictive diagnostic models have been developed to assist clinicians with RT-PCR results. However, the models do not provide the level of accuracy that is possible. To solve this problem, the proposed system designed a Dynamic Scale factor based Elephant Herding Optimization (DSEHO) with Weighted Support Vector Machine (WSVM) approach for covid-19 prediction. The initial analysis includes a dataset of COVID-19 patients. The data should be normalized using a min-max method. Independent Component Analysis (ICA) is introduced to decrease the dimensions. To select the optimal features, a Dynamic Scale factor based Elephant Herding Optimization (DSEHO) is implemented. For the prediction of COVID-19, Weighed Support Vector Machines (WSVM) is used. The proposed system provides improved accuracy, precision, recall, and F-measure over the previous system.