Wednesday June 23, 2010 at 5.00pm - Lecture theatre SMB.0.14, Stewart Mason Building
Jam Tomorrow – a view of Computational Intelligence
Professor Chris Hinde, Department of Computer Science
Computational intelligence is concerned with three major areas: Fuzzy and Vague Reasoning, Neural Networks and Evolutionary Systems.
Work in Fuzzy Systems has involved reasoning with and about contradictions and inconsistencies. Contradictions drive knowledge forward and can be used to discover structure. “Jam tomorrow” is a bi-lingual oxymoron. Jam is Latin for now, and contradicts tomorrow.
Neural Networks are used in many applications; however, they are treated as black boxes, where their internal workings cannot be observed. They can be trained to model functions based only on data points, but can result in modelling the wrong underlying concept.
Evolutionary Systems are one of the big surprises in computing. Why would anyone want to evolve a solution to a problem? Computer –based evolutionary systems have proved to be fast and effective ways of solving many problems.
A small system developed to change seismic data into music is also explored. The system itself is not intelligent or creative, but it does allow someone who is intelligent and creative to produce musical pieces that react to their environment.