Forecasting of Monkeypox Cases in the World Using the ARIMA Model

Abstract

While the Covid-19 epidemic in the world was not over yet, the monkeypox epidemic started. The monkeypox virus spread to more than 59 countries in 4 months. Computer-aided forecasting models are needed to effectively control this spread. It has been seen in previous outbreaks that time-series models are effective in estimating the impact of the epidemic and taking necessary precautions. In this study, different Automatic Regressive Integrated Moving Average (ARIMA) models were developed to successfully forecast the number of monkeypox cases in the World. Daily confirmed monkeypox cases data from 07 May-12 July 2022 were used in the study. 07 May 2022-02 July data were used in the training of ARIMA models. The prediction performances of the models were tested with the data of 03 July-12 July 2022. According to the test results, the ARIMA(2,2,1) model with the lowest RMSE=483, MAE=410, and MAPE=4.82 was determined as the most successful model. It has been determined that the determined ARIMA model is in good agreement with the real values with an average error value of around 5%. The number of monkeypox cases for the next 7-day was forecasted using ARIMA(2,2,1) model. While the model predicts the number of monkeypox cases to be 15056 for 19 July 2022, the actual number of cases is 15032 proves the model's success. This is the first study to estimate the number of monkeypox cases using the ARIMA method, and the results show that the ARIMA model is a convenient method for estimating the number of monkeypox cases.