Discrete choice and machine learning: Thomson and Thompson?

Classification methods from machine learning are receiving a lot of attention in the transportation modelling community. This is motivated by the access to large databases, and to various success stories reported in this research community. Discrete choice models have been used for decades to model the behaviour of travellers, faced to various types of decisions. In this talk, we will compare the two approaches, and highlight their differences. In particular, we will identify good and less good practices in each community.