Special Session on:
"Recent Advances in Evolving Fuzzy Systems"


Organizers: Edwin Lughofer (FLLL - edwin.lughofer@jku.at),
                  Dimitar Filev(Ford Motor Company - dfilev@ford.com),
                  Plamen Angelov (Lancaster University - p.angelov@lancaster.ac.uk)
 

 

Scope

- Novel Incremental and Evolving Learning Algorithms for different Types of Fuzzy Systems:

    - Evolving Fuzzy Classifiers

    - Evolving Takagi-Sugeno-Kang Type Fuzzy Systems

    - Evolving Neuro-Fuzzy Approaches

- Issues regarding Robustness, Stability and Process-Safety in Evolving Fuzzy Systems

- Evolving Techniques to address "Concept Drift"

- Evolving Intelligent Systems

- On-line Techniques for dealing with Model Uncertainties and Interpretability Issues

- Real-World Applications of Evolving Fuzzy Systems (e.g. On-line Identification, On-line Modelling, On-line Fault Detection and Decision Support Systems and many more)

 

Aim of the Session

The aim of this special session is to bring researchers in the field of Evolving Fuzzy Systems from all over the world together,
to exchange their ideas and approaches, to discuss and to present latest results on this emerging field.
During the last decade the concept of evolving fuzzy systems established as a useful and necessary methodology to address the
problems of incremental learning, adaptation and evolution of fuzzy systems during on-line operation modes
(basically in industrial processes).

Opposed to a batch data-driven design of fuzzy systems, evolving fuzzy systems are able to automatically adapt themselves
to new operating conditions and system states and hence guarantee a higher process safety, especially for highly dynamic and time-variant systems.
For on-line processes, incrementality of the methods plays a central role as a batch re-training phase usually does not terminate in real-time.
Another major topic which can be addressed with evolving fuzzy systems is the building of models from huge
data bases which cannot be loaded at once into the memory and hence have to be loaded in sample-wise manner.
This requires an incremental learning engine which also supports the automatic evolution and pruning of structural
components in the fuzzy systems on demand.
 

Past Events

In the past, an international workshop was organized on this topic, i.e.
- the 2006 International Symposium on Evolving Fuzzy Systems, Ambleside, UK, 7-9 September, 2006,  proceedings published by IEEE Press,
- followed by a special session for the FUZZ-IEEE 2007 (http://www.fuzzieee2007.org/specialsessions2.php#S4) and
the
2009 IEEE Workshop on Evolving and Self-Developing Intelligent Systems, IEEE Symposium Series on Computation Intelligencein    Nashville, TN, USA. (http://www.ieee-ssci.org/index.php?q=node/6).

 

Important Dates

- Tenative title and list of authors to be sent to the Special Session organizers: November 28, 2008.

- Deadline for Papers Submission (according IFSA/EUSFLAT 2009 instructions): January 9, 2009.

- Notification of acceptance/rejection: March 2, 2009.

- Deadline for Camera-ready papers: April 3, 2009.

 

Organizers Contact Details

Edwin Lughofer
Email address: edwin.lughofer@jku.at
Fuzzy Logic Laboratorium Linz-Hagenberg
Softwarepark 21
A-4232 Hagenberg
Tel.: +43 (0)7236 / 3343 435
Fax: +43 (0)7236 / 3343 434
Official HP: http://www.flll.jku.at/people/staff/lughofer.html
 

Dimitar Filev
Email address: dfilev@ford.com
Ford Motor Company
1 American Road
Dearborn, MI 48126-2798


Plamen Angelov
Email address: p.angelov@lancaster.ac.uk
Department of Communication Systems
Lancaster University
South Drive
Lancaster LA1 4WA, UK
phone: +44 (1524) 5-10391
fax: +44 (1524) 510489
Official HP: www.lancs.ac.uk/staff/angelov