Steve works on optimisation: making decisions so that things work better.  This can involve many technologies: constraint/integer/dynamic/stochastic/robust/bilevel programming, Boolean satisfiability, evolution and neuroevolution, reinforcement learning, and anything else that works (often hybrid methods).  Steve also works on aspects of machine learning and data analytics: feature selection, feature learning, deep learning and clustering.  He also works on combining optimisation and machine learning to solve large-scale problems from industry.