2021 · In plain language
How much of a population does a simulation need?
A review and case study of population downscaling in multi-agent transport simulation.
01
The question
Simulating a sample of a population can reduce computing requirements. The practical question is how that choice changes the behavior and results of the model.
02
The approach
The paper combines a review of population downscaling with a case study. It examines the trade-off between a smaller simulated population and the representation of the transport system.
03
The practical takeaway
Population sample size is a modeling decision that needs validation for the intended analysis. The paper is a starting point for examining that choice; a sample that works for one model should be checked before it is reused in another.