I work in probabilistic inference - the mathematical theory of acquisition of knowledge from data. While I am generally interested in all its aspects, the main focus of my work are machines that actively seek knowledge. Related scientific communities include machine learning, statistics, numerical optimisation, experimental design, control, operations research, stochastic analysis, numerical analysis, and robotics. I have particular expertise with approximate inference methods, such as free energy formulations and Monte Carlo methods, as well as kernel methods and nonparametric Bayesian methods. My work has found application in computer linguistics, cancer therapy, automated game playing, psychometrics, robotics, and others.

Further information, including a list of publications, can be found in my CV.