About me

I am working as a research software development engineer (RSDE) in the Machine Learning and Perception Group (MLP) at Microsoft Research in Cambridge, UK, in close collaboration with Prof. Chris Bishop. Projects from the last few years include

• improvements to the ink parser in the Tablet PC version of Windows;
• variational methods applied to hierarchical mixtures of experts and robust mixture models.

More recently, I have been working on figures (Matlab®, MetaPost), typesetting (LaTeX), exercise solutions and the errata for Chris Bishop's book, Pattern Recognition and Machine Learning.

Brief biography

I studied Computer Science at Linköping University, Sweden, 1990–1993. In autumn 1993, I went to Aston University, Birmingham, UK, for my fourth and final year. There I also did my MSc project, supervised by Prof. Chris Bishop in the Neural Computing Research Group. He convinced me to continue with a PhD, which I started in January 1995.

My PhD thesis was concerned with a latent variable model, called the generative topographic mapping (GTM), for capturing and visualizing non-linear low-dimensional structure in high-dimensional data. There is still a web page about the GTM, with a Matlab implementation amongst other things, but since I haven't worked on this since 1998, it is not up to date with what has happened in this field.

In February 1998, I moved to the Max-Planck-Institute of Cognitive Neuroscience in Leipzig Germany, where I spent my first two years working on analysis of functional magnetic resonance imaging (fMRI) data, using methods from machine learning. The following two years I worked as a programmer within an EU-funded project called SimBio, working on software for visualizing various forms of 3D medical data.

I moved to Microsoft Research in September 2002.