Evaluating the Impact of Traveling on COVID-19 Prevalence and Predicting the New Confirmed Cases According to the Travel Rate Using Machine Learning: A Case Study in Iran

COVID-19 which has spread in Iran from February 19, 2020, infected 1,550,142 people and killed 59,264 people until February 18, 2021. The immediate suggested solution to prevent the spread of this virus was to avoid traveling around. Unfortunately, on many occasions, this restriction was not enforced or respected by the citizens. Thus, the goal of this study was to evaluate the impact of traveling on the COVID-19 prevalence by measuring the correlation between traveling data and new confirmed cases of COVID-19 in Iran. The data consists of the daily traffic between Iran’s provinces, air traffic, and daily COVID-19 new confirmed cases. In the first step, the importance analysis was used to determine the impact of different kinds of traveling on the COVID-19 spread. In the second step, KNN, Random Forest (RF), and Support Vector Regression (SVR) were used to predict the effect of traveling on the number of new COVID-19 cases. Although the available data was very coarse and there were no details of inner-cities commute, an R-squared of 0.89 and 0.86 on the train and test datasets was achieved respectively, showing a positive correlation between the number of travels between states and the new confirmed cases of COVID-19. It was also shown that there was an 8-day incubation period. Consequently, by considering this period the voting regressor model reached 0.92 and 0.98 R
²
scores for test and train datasets respectively. The result confirms that one of the best ways to avoid the spread of the virus is limiting or eliminating traveling around.