Bayesian networks for constrained location choice modeling using structural restrictions and model averaging
Abstract
In this work, we propose a Bayesian network approach by using structural restrictions and a model averaging algorithm for modeling the location choice of discretionary activities. In a first stage, we delimit individuals’ location choice which is set by generating an ellipse that uses empirical detour factors and a home-work axis. The choice set is further refined by an individual’s space-time constraints in order to identify the constrained destination choice set. We use structural restrictions and a model averaging method to learn the network structure of the Bayesian network in order to predict the heuristics of individuals’ location selection. The empirical study shows the proposed method can effectively obtain Bayesian networks with a consistent dependency structure. The empirical study suggests activity schedule factors significantly influence location choice decisions.
EJTIR is licensed under a Creative Commons Attribution 4.0 International (CC BY 4.0) licence. The license means that anyone is free to share (to copy, distribute, and transmit the work), to remix (to adapt the work) under the following conditions:
- The original authors must be given credit
- For any reuse or distribution, it must be made clear to others what the license terms of this work are
- Any of these conditions can be waived if the copyright holders give permission
- Nothing in this license impairs or restricts the author's moral rights