On the use of ML "Good Practices" for DCMs
Responsable(s) :
Gael Lederrey, Michel Bierlaire, Nicholas Molyneaux
Description :
Machine Learning does not exist without Cross-Validation. This good-practice has been used for many years and has shown its utility. On the other hand, Discrete Choice Models are trained without Cross-Validation. In this project, we want to investigate if using Cross-Validation would really help Discrete Choice Models on the Forecasting side while keeping its behavioural component. The first phase of the project will be to implement the model proposed by Bierlaire et al. (2001) on the Swissmetro data using Larch. The second phase will be to implement a wrapper with Cross-Validation to train the same model and analyse the results. Finally, the third phase will be about finding other models in the literature, implement them, and analyse the results.
Collaboration with:
Type :
semester project
Pré-requis :
A good knowledge of Python is required. In addition, some general knowledge in Machine Learning and Discrete Choice Models (Course Mathematical Modelling of Behaviour) are highly recommended.
Submitted on :
September 14, 2018