Predicting the Effect of COVID-19 on Physical Activity of Survivors Using GSO and Hybrid Intelligent Model

Mild COVID-19 infection may take a long time to recover, especially when trying to return to exercise. Patients are left unsure of how, when, and if they should resume physical activity after COVID-19. Some people may have tried to resume normal exercise but were unable to, causing concern about returning to normal. In this work, a local dataset (questionnaire) was collected on recovered people, including age, gender, medical history, symptoms associated with infection, and information about their current physical activity. The data set was pre-processed and then entered into the feature selection algorithm (GSO) to get the best characteristics and then predict the impact of physical activity for people recovering from COVID-19 using two hybrid models: RF with the DT and RF with the LR. The results of performance measures showed that the hybrid model overcome on the traditional algorithms. This prediction helps people be more aware of the potential effects of their physical activity, so that they can be monitored and their health needs met quickly.