Route choice set generation using variational autoencoders

Abstract

Choice set generation is a challenging task, since the consideration set is generally unknown to the modelers, and the full choice set cannot be enumerated in real size networks. The proposed variational autoencoder approach (VAE) is motivated by the idea that the chosen alternatives must belong to the consideration set. The VAE approach explicitly considers maximizing the likelihood of including the chosen alternatives in the choice set, and infers the underlying generation process. The VAE approach for route choice set generation is exemplified using a real dataset. VAE-CNL model has the best performance in terms of goodness-of-fit and prediction performance, compared to models estimated with conventionally generated choice sets.