Intelligent judgements over health risks in a spatial agent-based model

Area of Science:

• Computational epidemiology
• Behavioral science
• Machine learning

Background:

• Infectious diseases pose a global threat, necessitating better predictive models.
• Current agent-based models (ABMs) often overlook the crucial role of risk perception in disease spread.
• Understanding how individuals perceive risk and adapt behavior is key to controlling epidemics.

Purpose of the Study:

• To develop an innovative agent-based model (ABM) integrating machine learning (ML) for behavioral dynamics.
• To simulate the impact of risk perception and adaptive behaviors on disease transmission.
• To compare disease spread patterns between intelligent and non-intelligent agents.

Main Methods:

• Developed a spatial agent-based model (ABM) incorporating Protection Motivation Theory.
• Integrated two Bayesian Networks (BNs) into a NetLogo-based Cholera ABM: BN1 for risk perception, BN2 for risk and coping behavior.
• Conducted computational experiments comparing zero-intelligent agents with agents exhibiting BN1 and BN2 intelligence.

Main Results:

• Simulated epidemic curves, risk perception dynamics, and coping strategy distributions for different agent intelligence levels.
• Demonstrated significant differences in disease spread patterns based on agent behavior.
• Quantified the impact of intelligent decision-making on disease incidence.

Conclusions:

• Integrating behavioral decision-making using ML into spatial ABMs is crucial for accurate disease modeling.
• This approach enhances the study of intervention strategies and the cumulative effects of behavioral changes.
• The findings highlight the importance of human behavior in infectious disease dynamics.