34

A talk by Michael Haley, <a href="http://www.bp.com/home.do?categoryId=1">BP International</a>, Chief Technology Office

<p>A talk by Michael Haley, <a href="http://www.bp.com/home.do?categoryId=1">BP International</a>, Chief Technology Office</p> 2006-02-09 13:00:00 Thursday Thu 09 February Feb 02 2006 13 00 2006-02-09 14:00:00 Thursday Thu 09 February Feb 02 2006 14 00 Thursday 09 February 2006, 1300-1400 0 0 C60 InfoLab21 2 Talk, Lecture or Seminar Talks, Lectures and Seminars <p>An opportunity to hear about the latest research by invited experts in their field. All are welcome to attend.</p> 33

A talk by <a href="http://public.rz.fh-wolfenbuettel.de/%7Eklawonn">Professor Frank Klawonn</a>, University of Applied Sciences, Braunschweig, Germany

<p>A talk by <a href="http://public.rz.fh-wolfenbuettel.de/%7Eklawonn">Professor Frank Klawonn</a>, University of Applied Sciences, Braunschweig, Germany</p> 2006-04-05 13:00:00 Wednesday Wed 05 April Apr 04 2006 13 00 2006-04-05 14:00:00 Wednesday Wed 05 April Apr 04 2006 14 00 Wednesday 05 April 2006, 1300-1400 0 0 C60 InfoLab21 2 Talk, Lecture or Seminar Talks, Lectures and Seminars <p>An opportunity to hear about the latest research by invited experts in their field. All are welcome to attend.</p> 32

By <a href="http://www.cs.ugal.ro/%7Ebocanialac/english/index.html">Dr. Cosmin Danut Bocaniala</a>

<p>By <a href="http://www.cs.ugal.ro/%7Ebocanialac/english/index.html">Dr. Cosmin Danut Bocaniala</a></p> <p>The talk focuses on a novel distributed methodology for complex causal systems. Here, a complex system represents a system whose global behaviour, which emerges from the interactions between its usually large number of basic components, is difficult to be accurately described via state-of-the-art modeling techniques. The methodology imitates the manner humans perform reasoning when confronted with mastering complex systems. The monitored complex system is partitioned into minimally separated and causally independent parts that can be separately modeled and understood.</p> <p>The fact that each region is causally independent from the rest of the model, allows allocating a dedicated local agent to each region, that will be able to perform the required task at local level. However, the local agents exchange information between them via partition's borders. The methodology provides minimal complexity for the communication process between agents. The research performed so far indicates that the methodology can be successfully applied for a variety of purposes on complex systems, which feature causal relationships between their basic components. During presentation, application examples from Fault Diagnosis, Particle Swarm Optimization and Autonomous Agents areas will be provided.</p> 2006-05-16 13:00:00 Tuesday Tue 16 May May 05 2006 13 00 2006-06-16 14:00:00 Friday Fri 16 June Jun 06 2006 14 00 Tue 16 May - Fri 16 June 2006 0 0 C60 InfoLab21 2 Talk, Lecture or Seminar Talks, Lectures and Seminars <p>An opportunity to hear about the latest research by invited experts in their field. All are welcome to attend.</p> 30

A talk by Professor Jim Kyle, Bristol University

<p>A talk by Professor Jim Kyle, Bristol University</p> 2006-06-08 13:00:00 Thursday Thu 08 June Jun 06 2006 13 00 2006-06-08 14:00:00 Thursday Thu 08 June Jun 06 2006 14 00 Thursday 08 June 2006, 1300-1400 0 0 C60 InfoLab21 2 Talk, Lecture or Seminar Talks, Lectures and Seminars <p>An opportunity to hear about the latest research by invited experts in their field. All are welcome to attend.</p> 29

A talk by Professor Les Barclay

<p>A talk by Professor Les Barclay</p> 2006-06-20 12:00:00 Tuesday Tue 20 June Jun 06 2006 12 00 2006-06-20 13:00:00 Tuesday Tue 20 June Jun 06 2006 13 00 Tuesday 20 June 2006, 1200-1300 0 0 C60 InfoLab21 2 Talk, Lecture or Seminar Talks, Lectures and Seminars <p>An opportunity to hear about the latest research by invited experts in their field. All are welcome to attend.</p> 28

By Dr. Gary Abel, British Antarctic Survey

<p>By Dr. Gary Abel, British Antarctic Survey</p> <p>There is a growing body of work that shows scale free behaviour in a wide range of phenomena in the Magnetosphere-Ionosphere system. These observations include the AE indices, magnetic fields, electric fields, auroral brightenings, convection flows and magnetic structures associated with reconnection. They have been analysed using a variety of techniques including spectral analysis, threshold crossings and structure function analysis. This seminar presents some analysis of convection velocities measured by the SuperDARN radar which show scale free complex behaviour. These results will be discussed in the wider context along with some of the implications of scale free behaviour in the Magnetosphere. These implications affect both the way we view the system and the framework within which we should view observations.</p> 2006-07-06 11:00:00 Thursday Thu 06 July Jul 07 2006 11 00 2006-07-06 12:00:00 Thursday Thu 06 July Jul 07 2006 12 00 Thursday 06 July 2006, 1100-1200 0 0 C60 InfoLab21 2 Talk, Lecture or Seminar Talks, Lectures and Seminars <p>An opportunity to hear about the latest research by invited experts in their field. All are welcome to attend.</p> 27

By Dr Edwin Lughofer, Johan Kepler University, Linz, Austria

<p>By Dr Edwin Lughofer, Johan Kepler University, Linz, Austria</p> <p>In nowadays industrial systems there are quite a lot requirements for an incremental modelling of system behaviours from data. These requirements arise due to a fast tracking of dependencies between system variables with online recorded measurements or due to huge data-bases which cannot be loaded into virtual memory at once. Improving the process security by extending already available models to new operating conditions or by adjusting them on the basis of some feedback from operators is an important issue as well.</p> <p>In this talk algorithms for a data-driven incremental learning of fuzzy models are demonstrated. The main focus is given to fuzzy basis function networks (as a specific type of Takagi-Sugeno fuzzy models), but some aspects how to extend the approach to other types of fuzzy models are illuminated as well. A modified version of vector quantization is exploited for rule evolution and an incremental learning of the rules' premise parts. The modifications include the generation of new clusters due to the nature, distribution and quality of new data, point-wise update of already existing clusters and an alternative distance strategy for selecting the most adjacent clusters for each new incoming sample. The premise part learning is connected in a stable manner with a recursive learning of rule consequent functions possessing linear parameters. Stability can be achieved by preventing the 'unlearning' effect in case of steady states and by ensuring that a suboptimal solution of the parameters can be achieved which is close to the optimal one in the least squares sense (i.e. with respect to least squares as underlying optimization function). Some aspects about maintaining the interpretability of the trained fuzzy models round off the algorithmic part of the talk. The second part is focussed on applications of data-driven evolving fuzzy models in industrial systems. These applications include online identification for high-dimensional static and dynamic data, online fault detection, refinement of expert-knowledge based fuzzy systems with data and model-based (inverse and predictive) control. Some evaluation results for online identification and fault detection are mentioned.</p> 2006-09-12 14:00:00 Tuesday Tue 12 September Sep 09 2006 14 00 2006-09-12 15:00:00 Tuesday Tue 12 September Sep 09 2006 15 00 Tuesday 12 September 2006, 1400-1500 0 0 C60 InfoLab21 2 Talk, Lecture or Seminar Talks, Lectures and Seminars <p>An opportunity to hear about the latest research by invited experts in their field. All are welcome to attend.</p> 26

A talk by Dr Mike Warrington, Department of Engineering, University of Leicester

<p>A talk by Dr Mike Warrington, Department of Engineering, University of Leicester</p> 2006-09-19 13:00:00 Tuesday Tue 19 September Sep 09 2006 13 00 2006-09-19 14:00:00 Tuesday Tue 19 September Sep 09 2006 14 00 Tuesday 19 September 2006, 1300-1400 0 0 C60 InfoLab21 2 Talk, Lecture or Seminar Talks, Lectures and Seminars <p>An opportunity to hear about the latest research by invited experts in their field. All are welcome to attend.</p> 25

By Professor Maria Petrou, Imperial College, London

<p>By Professor Maria Petrou, Imperial College, London</p> <p>The human vision system has been the inspiration for many computer vision approaches. However, the human vision system has its limitations.</p> <p>The talk will concentrate on methods used to allow us to see characteristics in the data that are not obvious to the human eye. It will show examples of 3D texture analysis for MRI data and also results of tumour boundary detection using high order statistics.</p> 2006-10-26 14:00:00 Thursday Thu 26 October Oct 10 2006 14 00 2006-10-26 15:00:00 Thursday Thu 26 October Oct 10 2006 15 00 Thursday 26 October 2006, 1400-1500 0 0 C60 InfoLab21 2 Talk, Lecture or Seminar Talks, Lectures and Seminars <p>An opportunity to hear about the latest research by invited experts in their field. All are welcome to attend.</p> 24

A talk by Dr George Magoulas, Birkbeck College, University of London

<p>A talk by Dr George Magoulas, Birkbeck College, University of London</p> 2006-12-06 13:00:00 Wednesday Wed 06 December Dec 12 2006 13 00 2006-12-06 14:00:00 Wednesday Wed 06 December Dec 12 2006 14 00 Wednesday 06 December 2006, 1300-1400 0 0 C60 InfoLab21 2 Talk, Lecture or Seminar Talks, Lectures and Seminars <p>An opportunity to hear about the latest research by invited experts in their field. All are welcome to attend.</p>