2013 Academic Year Seminars
Substantial quantities of data are being generated within the biomedical sciences and the successful integration of different types of data remains an important challenge. We begin the talk with an overview of some of our lines of investigation in this context. We start with supervised learning and outline several approaches to multi-kernel learning we have developed which can handle disparate types of input data. These methods can give state-of-the-art performance on certain protein fold prediction tasks, for example. We also briefly review work on the joint unsupervised modeling of two types of data believed functionally linked such as microRNA and gene expression array data from the same cancer patient. Finally, to extend the range of kernels which can be used for supervised learning, we outline contemporary work based on devising proxy kernels for handling alignment scores.
Dr. Colin Campbell
Dr. Campbell gained his first degree from Imperial College, London and his PhD from the Department of Mathematics, King's College, London. He is currently a Reader in the Department of Engineering Mathematics, University of Bristol. His research interests cover machine learning, including probabilistic graphical models and kernel-based methods, algorithm design and the applications of machine learning techniques in bioinformatics, cancer bioinformatics and medical informatics. His research is currently supported by EPSRC, Cancer Research UK and PASCAL2.