Передовые методы в биологии и медицине
Открытый доступ

Абстрактный

A Comparative Study of ARIMA and SARIMA Models to Forecast Lockdowns due to SARS-CoV-2

Hardik Chhabra*

By predicting the implementation of intelligent lockdown before the onset of waves, this research offers an efficient method to address the COVID-19 pandemic. This study introduces unified Autoregressive Integrated Moving Average (ARIMA) and Seasonal Autoregressive Integrated Moving Average (SARIMA) models that are capable of predicting lockdowns in more than 200+ nations. The proposed model was trained over 18,000 dataset of 237 nations and has a response time of 2.5 months. The auto-ARIMA model was used to pick the initial variations of the model parameters and then the optimal model parameters were found based on the best match between the forecasts and test data. The models reliability was evaluated using the analytical methods Auto Correlation Function (ACF), Partial Auto-Correlation Function (PACF), Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC). These models are trained using data acquired from World Health Organizations data repository. The two ARIMA and SARIMA models are clearly gaining an advantage over other studies by having a rapid response time. Besides, a brief comparison of trained ARIMA and SARIMA models are presented and the ARIMA model gained an upper hand due to its accuracy. Additionally, the models are able to predict confirmed death and confirmed cases of COVID. This research shows to be highly beneficial for decision-making about the implementation of smart-lockdowns and could provide another dimension to time-series analysis, which is strongly dependent on models having better response time.