Professor Williamson has developed scientific theory and widely used practical algorithms to solve machine learning problems. His best known work is in the field of “kernel machines”, a particular form of machine learning methods based on the geometry of infinite dimensional spaces. In addition to developing new powerful theoretical frameworks to analyse such techniques (the fact they are effectively working in infinite dimensions causes difficulties) he developed three widely used practical algorithms – the “nu” Support Vector Machine, the one-class SVM, and a simple online SVM. These are popular because they are effective, efficient and the adjustable parameters are readily interpretable.