A theoretical framework for modeling activity-travel scheduling decisions in non-stationary environments under conditions of uncertainty and learning

This paper describes a Bayesian framework for modelling dynamic activity-travel choice under conditions of uncertainty and learning. The approach proposed differs in several respects from earlier approaches. First, the decisions considered are framed as re-scheduling decisions and, hence, the model takes the impact of a full actvity. schedule into account. Second, the model is based on the assumption that individuals' perceptions of uncertainty are represented as probability distributions across possible states of the (transport) system and that belief updating and structural learning can be modeled using Bayesian methods. The paper outlines the framework and discusses ways to specify and test the models in future research.