Home

Nowadays, AI is used in almost every corner of society. For a long time, choice behaviour analysis was considered the exclusive domain of theory-driven (parametric) methods, in particular discrete choice models (founded in micro-economic theory). Recently this has changed. Thrilling new AI methods are being developed to obtain deeper behavioural insights on choice behaviour. These methods can complement and extend the theory-driven tools of choice behaviour researchers as they typically are: 

(1) More flexible & (2) Less restrictive in terms of data types.

Researchers at AI4ChoiceLab aim to bridge the gap between data-driven AI methods and theory-driven discrete choice methods, for better understanding human choice behaviour. 

 
software-300x300.jpg

SOFTWARE

Exploring new methods may be a time consuming practice, especially when unfamiliar with a field’s tacit practices. Therefore, we provide software code for the methods I have developed in our papers. 

electric-light-bulb_1f4a1.png

GLOSSARY OF CONCEPTS

This glossary lists relevant concepts in the AI literature, explained in an accessible way for researchers with a general background in statistics. 

Overview_of_literature.jpg

OVERVIEW OF EMERGING LITERATURE

Overview of emerging literature on AI for choice behaviour research. 

Data_driven_vs_theory_driven_methods_v2.

DATA-DRIVEN VS THEORY-DRIVEN MODELS OF CHOICE BEHAVIOUR

What is the difference between data-driven and theory-driven models? Here, we explain the core differences and similarities of the two powerful modelling paradigms for choice modellers.  

Translation_table.jpg

CONCEPT TRANSLATION TABLE

Choice behaviour researchers may feel lost in translation when browsing the AI literature. But, many concepts are shared across fields. This table helps researchers to navigate the AI literature.