KYC-Analytics was focused on developing foundations and techniques for next generation KYC predictive analytics under increasing data heterogeneity and volatility of the customer behavior.
In collaboration with the compliance department of Rabobank we have developed a series of sub-projects on the cutting edges of text-mining, visual analytics, deep learning, concept drift and heterogeneous evolving data handling. One of the cross-cutting themes included anomaly detection and post-processing of alerts with human-in-the-loop, e.g. in the application of fraud detection in transactional, graph, and text data. In many of the sub-projects also the issues of ethics-awareness, transparency and accountability of predictive analytics have been addressed.