Calibrating a Stochastic, Agent-Based Model Using Quantile-Based Emulation

Arindam Fadikar, Dave Higdon, Jiangzhuo Chen, Bryan Lewis, Srinivasan Venkatramanan, and Madhav Marathe

SIAM/ASA Journal on Uncertainty Quantification, 6(4):1685-1706, 2018

SIAM

Quantile-based emulation for stochastic agent-based model calibration

Abstract

This paper develops a quantile-based emulation strategy for calibrating stochastic agent-based models where output variability is substantial and distributional behavior matters.

The approach focuses on emulating informative output quantiles rather than only means, enabling robust parameter calibration under heteroskedastic simulation uncertainty.

The framework demonstrates improved calibration performance for complex stochastic simulators and provides a practical pathway for uncertainty-aware inference in agent-based epidemiological modeling.

Citation

BibTeX

@article{fadikar2018quantile,
  title={Calibrating a Stochastic, Agent-Based Model Using Quantile-Based Emulation},
  author={Fadikar, Arindam and Higdon, Dave and Chen, Jiangzhuo and Lewis, Bryan and Venkatramanan, Srinivasan and Marathe, Madhav},
  journal={SIAM/ASA Journal on Uncertainty Quantification},
  volume={6},
  number={4},
  pages={1685--1706},
  year={2018},
  publisher={SIAM},
  doi={10.1137/17M1161233}
}