Forecasting the COVID-19 Pandemic in Saudi Arabia Using a Modified Singular Spectrum Analysis Approach: Model...

Area of Science:

• Epidemiology
• Data Science
• Time Series Analysis

Background:

• Global infectious diseases pose significant health threats worldwide.
• The COVID-19 pandemic, caused by SARS-CoV-2, has become a major global health crisis.
• Accurate forecasting of COVID-19 spread is crucial for public health management.

Purpose of the Study:

• To develop a reliable and interpretable model for COVID-19 case forecasting in Saudi Arabia.
• To accurately predict the peak of the COVID-19 pandemic in Saudi Arabia.
• To analyze, decompose, and forecast confirmed COVID-19 cases using a novel approach.

Main Methods:

• A modified singular spectrum analysis (SSA) approach was employed for COVID-19 data analysis.
• The method was enhanced for improved separability, signal extraction, and noise reduction.
• Vector SSA was utilized for predicting future data points and the pandemic peak.

Main Results:

• The modified SSA identified two key eigenvalues (r=2) for signal separation from noise.
• Forecasting indicated the COVID-19 peak in Saudi Arabia around May-June 2020.
• The pandemic was predicted to conclude between late June and mid-August 2020, with approximately 330,000 total cases.

Conclusions:

• Modified SSA demonstrates effective performance in analyzing noisy time series data for COVID-19.
• The method reliably identifies signal components and predicts future trends.
• Vector SSA provides accurate forecasting of confirmed cases and pandemic peaks.