Methodology and a case study of statistical agent-based modelling of discrete choices: Comparison and integration with econometrics models

摘要

This project starts with an econometric analysis with two probabilistic equation-based models to analyse a real world problem - the effect of a road user charging scheme in the Upper Derwent Valley. Following this, a computer simulation model is developed to overcome the disadvantages of the equation-based models. The computer simulation model integrates the two equation-based models, a Markov queue network model, and a Minority Game. In other words, this simulation model is partially statistical and partially agent-based, and is termed a statistical agent-based model. The results of the two analyses are presented with the main focus being on the process of simulation and methodological aspects. The statistical agent-based model successfully integrates models from different disciplinary backgrounds, and shows interesting processes of simulation.