Special Session on "Evolutionary Computation for Knowledge Discovery
(ECKD)"
2010 IEEE World Congress on Computational Intelligence (IEEE WCCI 2010)
2010 IEEE Congress on Evolutionary Computation (IEEE CEC 2010)
http://www.wcci2010.org/
July 18-23, Barcelona, Spain
Submission Deadline: January 31st, 2010
Organizers:
Jaume Bacardit, University of Nottingham, UK
Jorge Casillas, University of Granada, Spain
Evolutionary Computation for Knowledge Discovery (ECKD) is a broad term
than includes all kinds of usage of evolutionary computation algorithms
to automatically derive abstractions from data sets.
Traditionally, most of these systems have been rule-based, being
learning classifier systems one of the best examples of this. However,
through the years we have seen an explosion of knowledge representations
such as fuzzy rules, trees, hyperspheres, neural representations,
kernels, etc. Large progress has also been made in several other
directions, such as tackling a broad range of learning problems, dealing
with multiple objectives, proposing theoretical models of the behavior
and performance of ECKD methods, or improving their scalability. Through
all these improvements, ECKD methods have the potential to become very
important players in the Data Mining and Knowledge Discovery field due
to their robustness, flexibility, scalability, and interpretability
capacity.
The motivation behind this special session is to provide a forum for
active ECKD researchers to present and discuss recent ideas and results,
propose and discuss current and future directions of ECKD research, and
expose the work on ECKD to the wider evolutionary computation and
machine learning communities. Collaborations between different ECKD
research subfields such as learning classifier systems and genetic fuzzy
systems are specially welcomed.
Topics welcomed in this special session include (but are not limited to)
the following:
- ECKD by novel evolutionary (and related) learning/optimization mechanisms
* Genetic algorithms
* Genetic programming
* Estimation of distribution algorithms
* Ant colony optimization
* Particle swarm optimization
* Memetic algorithms
- ECKD with advanced knowledge representations
* Fuzzy and non-fuzzy rule-based systems
* Tree representations
* Kernel representations
* Neural representations
* Ensemble representations
* Hybridization with other non-evolutionary techniques
- ECKD for challenging tasks
* Supervised learning (classification, regression, ...)
* Unsupervised learning (clustering, association rules, ...)
* Hierarchical/ordinal classification
* Feature/instance selection
* Data streams
- Efficiency-enhancement methods in ECKD
* Parallel architectures and distributed computing
* Graphics processing units
* Windowing techniques
* Fitness surrogates
- Theoretical advances in ECKD
- Real-world applications of ECKD
The papers must be submitted through the conference website
(http://www.wcci2010.org/) choosing this special session (ID S087) from
the available list when required. Formatting instructions are available
at http://www.wcci2010.org/submission.
Important dates:
- Paper submission: January 31, 2010
- Notification of paper acceptance: March 15, 2010
- Final paper submission: May 2, 2010
- IEEE WCCI 2010 Conference: July 19, 2010
--
-------------------------------------------------------------------
Jaume Bacardit, PhD
Lecturer in Bioinformatics
University of Nottingham
Automated Scheduling, Planning and Optimisation research group,
School of Computer Science, Jubilee Campus, Nottingham, NG8 1BB, UK
Multidisciplinary Centre for Integrative Biology,
School of Biosciences, Sutton Bonington, LE12 5RD, UK
Tel: +441159516276
Fax: +44 1159516292
Email: jaume _dot_ bacardit _at_ nottingham _dot_ ac _dot_ uk
Web: http://www.cs.nott.ac.uk/~jqb
--------------------------------------------------------------------
|