[Project in collaboration with Telenor-NTNU AILab]
In this project we wish to explore the application of deep neural networks to the classification of nodes in a network of money transactions. We have a fairly rich data set of real transactions; and the ultimate goal is to find a meaningful way to define distinct roles for the nodes in this transaction network. These roles will then tell us something about the real world from which the data are taken. We are primarily interested in trying unsupervised classification; but some degree of supervision may also be possible to aid the role discovery process. To our knowledge, this is a novel application of deep learning – and we are open to the project having an explorative component.
(Project in collaboration with Telenor Research)