Outbreak minimization v.s. influence maximization: an optimization framework

Area of Science:

• Epidemiology
• Network Science
• Computational Biology

Background:

• Targeted immunization identifies
• super spreaders
• to contain epidemic outbreaks (e.g., COVID-19) by minimizing disease spread over human contact networks.

Purpose of the Study:

• Establish the equivalence between targeted immunization and the influence maximization problem.
• Develop an effective methodology for targeted immunization using influence maximization principles.

Main Methods:

• Formulated the targeted immunization problem and demonstrated its equivalence to influence maximization under the Linear Threshold diffusion model.
• Developed a Benders' decomposition algorithm to solve the resulting optimization problem.

Main Results:

• Evaluated the optimization approach's performance and scalability on real-world large-scale networks.
• Demonstrated that the proposed approach yields more effective solutions compared to existing methods.

Conclusions:

• Confirmed the equivalence between outbreak minimization and influence maximization.
• Highlighted the importance of identifying optimal immunization groups for infectious disease containment.
• Proposed a methodology to aid policymakers in determining optimal immunization coverage levels.