Sensitivity analysis for agent-based models : a low complexity test-case

摘要

Methodologies for sensitivity analysis are considered to be of great importance for analyzing agent-based models (ABMs), even more because calibration and validation of ABMs often prove problematic. Different methodologies for sensitivity analysis may help to understand ABM dynamics and (thus) aid in the calibration and validation of ABMs. However, model complexity of ABMs is a significant hinderance for a detailed research on which (combination of) sensitivity analysis methods may provide the best option. We present here an agent-based model of low complexity to be used as test case for different methodologies of sensitivity analysis.