A Parallel Sliding Region Algorithm to Make Agent-Based Modeling Possible for a Large-Scale Simulation: Modeling...

Area of Science:

• Computational epidemiology
• Infectious disease modeling
• Agent-based modeling

Background:

• Agent-based models (ABMs) are valuable for simulating infectious disease epidemiology.
• Large-scale ABM simulations face significant computational cost limitations.

Purpose of the Study:

• To enhance the computational efficiency of large-scale ABM simulations.
• To introduce and evaluate a parallelizable sliding region algorithm (SRA) for ABM.

Main Methods:

• Developed a parallelizable sliding region algorithm (SRA) for ABM.
• Modeled hepatitis C epidemics using real demographic data from Saskatchewan, Canada.
• Compared SRA performance against a nonparallelizable ABM approach.

Main Results:

• The parallelizable SRA demonstrated significant computational time savings.
• SRA produced comparable results to nonparallelizable methods in province-wide simulations.
• The SRA approach is generalizable for country-wide simulations.

Conclusions:

• The developed SRA improves computational efficiency for large-scale ABM simulations.
• This parallel algorithm enables large-scale ABM with limited computational resources.
• SRA facilitates more accessible and scalable epidemiological modeling.