Bayesian Calibration of Stochastic Agent Based Model via Random Forest DOI
Connor Robertson, Cosmin Safta, Nicholson Collier

et al.

Statistics in Medicine, Journal Year: 2025, Volume and Issue: 44(6)

Published: March 13, 2025

ABSTRACT Agent‐based models (ABM) provide an excellent framework for modeling outbreaks and interventions in epidemiology by explicitly accounting diverse individual interactions environments. However, these are usually stochastic highly parametrized, requiring precise calibration predictive performance. When considering realistic numbers of agents properly stochasticity, this high‐dimensional can be computationally prohibitive. This paper presents a random forest‐based surrogate technique to accelerate the evaluation ABMs demonstrates its use calibrate epidemiological ABM named CityCOVID via Markov chain Monte Carlo (MCMC). The is first outlined context CityCOVID's quantities interest, namely hospitalizations deaths, exploring dimensionality reduction temporal decomposition with principal component analysis (PCA) sensitivity analysis. problem then presented, samples generated best match COVID‐19 hospitalization death Chicago from March June 2020. These results compared previous approximate Bayesian (IMABC) results, their performance analyzed, showing improved computation.

Language: Английский

Bayesian Calibration of Stochastic Agent Based Model via Random Forest DOI
Connor Robertson, Cosmin Safta, Nicholson Collier

et al.

Statistics in Medicine, Journal Year: 2025, Volume and Issue: 44(6)

Published: March 13, 2025

ABSTRACT Agent‐based models (ABM) provide an excellent framework for modeling outbreaks and interventions in epidemiology by explicitly accounting diverse individual interactions environments. However, these are usually stochastic highly parametrized, requiring precise calibration predictive performance. When considering realistic numbers of agents properly stochasticity, this high‐dimensional can be computationally prohibitive. This paper presents a random forest‐based surrogate technique to accelerate the evaluation ABMs demonstrates its use calibrate epidemiological ABM named CityCOVID via Markov chain Monte Carlo (MCMC). The is first outlined context CityCOVID's quantities interest, namely hospitalizations deaths, exploring dimensionality reduction temporal decomposition with principal component analysis (PCA) sensitivity analysis. problem then presented, samples generated best match COVID‐19 hospitalization death Chicago from March June 2020. These results compared previous approximate Bayesian (IMABC) results, their performance analyzed, showing improved computation.

Language: Английский

Citations

0