My research interests revolve mainly around machine learning, the application of adaptive statistical models to the problems of automated classification, prediction, planning and knowledge discovery. My work thus far has focused on problems in computational genomics, mainly concerned with a semi-supervised approach to gene function prediction. I'm currently also involved in a collaboration that attempts to predict cancer mortality from gene expression measurements.

Some of my key interests include:

• Unsupervised learning in neural networks (e.g. deep belief nets)
• Semi-supervised learning
• Computer vision, especially applications to microscope imaging
• Adaptive noise models for DNA microarrays
• The interface between machine learning and theoretical/computational neuroscience (i.e. models of how the brain works)