Assessing countries' performances against COVID-19 via WSIDEA and machine learning algorithms.

Area of Science:

• Public Health
• Data Science
• Epidemiology

Background:

• The COVID-19 pandemic has had a profound global impact, necessitating research into effective response strategies.
• Millions infected worldwide, driving urgent demand for solutions.

Purpose of the Study:

• To evaluate and rank the performance of 142 countries in managing the COVID-19 outbreak.
• To identify key parameters influencing national response effectiveness.

Main Methods:

• A novel three-stage framework integrating data envelopment analysis (DEA) and machine learning algorithms.
• Clustering using k-means and hierarchical methods.
• Efficiency analysis via weighted stochastic imprecise DEA, followed by decision tree and random forest parameter analysis.

Main Results:

• Identified three optimal clusters for country performance classification.
• 20 countries achieved full effectiveness; 36% reached 90% effectiveness.
• GDP, smoking rates, and diabetes prevalence did not significantly impact effectiveness levels.

Conclusions:

• The study provides a robust framework for assessing national pandemic response performance.
• Effectiveness is not directly correlated with economic factors or specific health indicators like smoking and diabetes.