This topic combines two of the most difficult or least understood ML techniques and challenges. Fraud detection invariably falls short of complete automatic detection because of the false positive rate and the need for at least some human intervention, typically on a case-by-case basis. Deep Learning, one of the most far flung borders of ML research utilizing neural net architecture but unsupervised model development.
Our friends at H2O University http://university.h2o.ai/ referred me to this very interesting 15 minute video by Venkatatesh Ramanathan who shows a real-world and very big data application of Deep Learning to fraud detection at Paypal.
See it here: