Advanced forecasting of COVID-19 epidemic: Leveraging ensemble models, advanced optimization, and decomposition...

Area of Science:

• Epidemiology
• Machine Learning
• Data Science

Background:

• Accurate forecasting of COVID-19 is critical for effective public health interventions.
• Previous epidemic modeling studies were limited by using single forecasting techniques.
• Data volatility in COVID-19 series presents a significant challenge for accurate prediction.

Purpose of the Study:

• To develop and evaluate a novel ensemble framework for improved COVID-19 forecasting.
• To integrate multiple machine learning methods for enhanced prediction accuracy.
• To address the limitations of single-model approaches in complex epidemic prediction.

Main Methods:

• An ensemble framework integrating kernel ridge regression (KRidge), Deep random vector functional link (dRVFL), and ridge regression (L-K_Ridge-d_RVFL-R_idge).
• Optimization using adaptive differential evolution and particle swarm optimization (A-DEPSO).
• Input variable decomposition via time-varying filter empirical mode decomposition (TVF-EMD) and feature selection using light gradient boosting machine (LGBM).

Main Results:

• The proposed ensemble model achieved high correlation coefficients (R) for COVID-19 forecasting in Italy (t+10 = 0.965, t+14 = 0.961) and Poland (t+10 = 0.952, t+14 = 0.940).
• The model demonstrated superior performance compared to other models in both case studies.
• Experimental results confirmed the model's outstanding performance in complex epidemic prediction scenarios.

Conclusions:

• The novel ensemble model offers exceptional accuracy and resilience in COVID-19 forecasting.
• The integrated approach outperforms existing models, providing a more reliable tool for epidemic prediction.
• This framework has significant potential for application in managing and controlling infectious disease outbreaks.