Forecasting COVID19 parameters using time-series: KSA, USA, Spain, and Brazil comparative case study

Area of Science:

• Epidemiology
• Data Science
• Public Health

Background:

• The COVID-19 pandemic necessitates accurate forecasting of cases and deaths for effective control strategies.
• Saudi Arabia (KSA) and other high-impact countries require reliable methods to predict disease spread.

Purpose of the Study:

• To forecast COVID-19 cases and deaths in KSA using time-series statistical techniques.
• To validate Exponential Trend Smoothing (ETS) and Drift models for case and death forecasting, respectively.
• To compare model performance against existing studies using key evaluation metrics.

Main Methods:

• Application of time-series analysis and statistical forecasting techniques, including Exponential Smoothing (SES, Holt, ETS) and Linear Regression (Drift).
• Forecasting of COVID-19 cases and deaths for KSA and major affected countries (US, Spain, Brazil).
• Validation of forecast results using four distinct evaluation measures, including Root Mean Square Error (RMSE).

Main Results:

• The Exponential Trend Smoothing (ETS) model demonstrated superior efficiency in forecasting COVID-19 cases.
• The Drift model was found to be effective for forecasting COVID-19 deaths.
• In KSA, the ETS model achieved an RMSE of 18.44, significantly outperforming state-of-the-art studies with an RMSE of 107.54.

Conclusions:

• The proposed ETS and Drift models offer a robust and accurate approach to forecasting COVID-19 pandemic parameters.
• These validated models can serve as a benchmark for pandemic management and control efforts globally.
• Accurate forecasting is essential for resource allocation and implementing timely public health interventions.