Paul Dunphy

Ismini Psychoula, Andreas Gutmann, Pradip Mainali et al.

The application of machine learning to support the processing of large datasets holds promise in many industries, including financial services. However, practical issues for the full adoption of machine learning remain with the focus being on understanding and being able to explain the decisions and predictions made by complex models. In this paper, we explore explainability methods in the domain of real-time fraud detection by investigating the selection of appropriate background datasets and runtime trade-offs on both supervised and unsupervised models.

Paul Dunphy, Fabien A. P. Petitcolas

The emergence of distributed ledger technology (DLT) based upon a blockchain data structure, has given rise to new approaches to identity management that aim to upend dominant approaches to providing and consuming digital identities. These new approaches to identity management (IdM) propose to enhance decentralisation, transparency and user control in transactions that involve identity information; but, given the historical challenge to design IdM, can these new DLT-based schemes deliver on their lofty goals? We introduce the emerging landscape of DLT-based IdM, and evaluate three representative proposals: uPort; ShoCard; and Sovrin; using the analytic lens of a seminal framework that characterises the nature of successful IdM schemes.