Bridging ML and behavioural models: An overview of deep learning strategies for choice modelling: current solutions and future directions

Beschrijving

There have been a number of innovative approaches in utilizing deep learning strategies for choice modelling in recent years. These strategies, initially developed for deep neural networks to solve various computational and complexity problems, are now being applied to discrete choice modelling. In this 'quick-fire' presentation, I will present a brief overview of recent works which have incorporated deep learning methods into choice models, and subsequently, identify future challenges and opportunities for deep learning in choice modelling research.

Periode	8 nov. 2021
Gehouden op	Alan Turing Institute, Verenigd Koninkrijk
Mate van erkenning	Internationaal