Prediction of COVID-19 case and death numbers in Turkey by using artificial intelligence techniques

The coronavirus, which emerged in China in the last months of 2019, soon turned into a pandemic that affected the whole world. The countries of the world, which were caught unprepared for such a pandemic, carried out data analyzes to be able to take correct and timely measures at that time by using artificial intelligence techniques to analyze the current situation and to prevent the handicaps arising in many areas due to the virus. In this study, using the Turkey coronavirus data set, predictive analysis was performed with three different boosting algorithms: AdaBoost, CatBoost, and XGBoost algorithms. In addition, predictions related to the considered data set were made by using ANN algorithms whose hyperparameters were optimized with the grid search algorithm. Basic performance metrics are used to evaluate and interpret the results of the analysis correctly. Necessary comparisons are made for four different analyzes on the same data set and the most appropriate algorithm is determined for the Turkey coronavirus data set.