Network-based analysis of national strategies for COVID-19 management

The COVID-19 pandemic highlighted the need for systematic evaluation of national response strategies that account for complex interdependencies across health, economic, and demographic dimensions. This study introduces a network-based framework that integrates machine learning for feature selection, entropy weighting for objective prioritization, and the Analytic Network Process to capture interdependencies among criteria and countries. Analysis of 147 countries shows that testing capacity, demographic structure, and healthcare resilience are decisive in shaping outcomes. High-performing nations, including Finland and Denmark, combined widespread testing with robust infrastructure and stringent public health measures, while resource-constrained countries faced significant challenges. Random Forest importance underscored the role of life expectancy and median age, while entropy weights emphasized testing and mortality-related indicators. Hierarchical clustering revealed regional performance patterns that align with socioeconomic resilience. The findings provide policymakers with a comprehensive tool to design evidence-based interventions, strengthen preparedness, and foster international cooperation in future health crises.