Matthew Studley

Lecturer
Department of Computing, Engineering and Mathematical Sciences
University of the West of England
Coldharbour Lane
Frenchay
Bristol
UK
BS16 1QY

Specialist Research Area
Machine learning, biologically-inspired robotics.

After a degree in Biology, and a career in industry including R&D in AI and telecoms, e-commerce consultancy and architecting a global unified messaging system, Matt Studley came to the UWE in 2001 to research for a PhD in Machine Learning and Robotics. His Thesis was on the use of Learning Classifier Systems - a form of machine lerning that uses elements of both Evolutionary Computing and Reinforcement Learning - in problems where more than one goal exist. In such problems, actions must be learned optimally, and the optimal balance between these actions must also be learned. An example of this might be a robot that needs to operate in dark places, but also return to the sunlight to recharge its batteries. Such optimally learned behaviour was demonstrated on a robotic platform and in simulation. After his PhD, Matt worked on a two year collaborative project funded by the EPSRC to deliver a suite of data mining tools on the National Supercomputer at Manchester, The aim of the project was to provide simple access to a datamining tool that would allow people with little knowledge of programming or data mining to discover useful facts about VERY LARGE datasets which would be impractical or intractable on desktop machines. After delivering the SCDM system, Matt started work as a Lecturer in Robotics. He looks forward to further research!