Improving the understanding and prediction of human behaviour over time.

点击翻译按钮获取中文摘要

Developing accurate behavioural models is a core challenge in many disciplines to inform decision and policy making. Considering this, diverse approaches have been implemented to model human behaviour according to the context. For instance, in the field of transport research, modelling techniques grounded in behavioural theories are estimated to capture the causal mechanisms between outcomes, but these methods do not fully exploit the richness of the data. In contrast, in mathematics, more complex and robust structures might have higher prediction capabilities but lack foundation in behavioural theories. In this context, the research aims to provide a better toolkit to evaluate behavioural outcomes over time by incorporating some of the mathematical methods to deal with longitudinal data into choice modelling frameworks typically used in transportation research, to specify high quality models with better forecasting capabilities.