Discovering optimal strategies for mitigating COVID-19 spread using machine learning: Experience from Asia

Area of Science:

• Epidemiology
• Data Science
• Public Health

Background:

• COVID-19 poses a significant global health challenge requiring data-driven interventions.
• Accurate forecasting of disease prevalence and mortality is crucial for effective pandemic management.

Purpose of the Study:

• To develop a spatiotemporal analysis framework for predicting COVID-19 spread.
• To identify key environmental and social factors influencing virus transmission.
• To optimize intervention strategies for pandemic containment.

Main Methods:

• Utilized an ensemble model combining random forest regression for prediction.
• Employed the Non-dominated Sorting Genetic Algorithm II (NSGA-II) for multi-objective optimization.
• Validated the framework across four Asian countries: Japan, South Korea, Pakistan, and Nepal.

Main Results:

• The random forest model accurately predicted daily confirmed cases and death rates.
• Temperature and social stringency index were identified as the most impactful features.
• NSGA-II found Pareto optimal solutions, potentially reducing infection and mortality by over 95%.

Conclusions:

• The developed framework provides valuable insights into disease evolution and forecasting.
• Social factors, particularly stringency index, offer greater potential for pandemic control than environmental factors.
• The findings support evidence-based decision-making for sustainable pandemic intervention and prevention.