NEURAL NETWORKS 2008
30h (lecture) + 30h (laboratory)
lectured by Anna Bartkowiak
Attendance:
Students with some primary knowledge on Numerical analysis,
Probability and Statistics, also Matlab programming
Topics:
Feedforward neural networks:
MLP - Multilayer perceptron,
RBF - Radial basis networks,
Data preprocessing and Visualization
Standardization
PCA - Principal components analysis
Canonical discriminant analysis
Recurrent: Elman network ?, Jordan network ?
Self-organizing networks:
SOM - self-organizing map (Kohonen)
Neural gas
GTM - Generative topographic mapping
Meta-analysis for decision making: Networks ensembles,
Network Committees, Stacked networks.
Laboratory -- is an essential part of the course.
Students obtain problems from the domain of pattern recognition,
like Classification and Clustering (medical diagnosis, similarity
of images, letter and face recognition, person authorization
from palm scans).
We will use the following software in Matlab:
Netlab (free, Nabney), Som Toolbox (free, HUT, Helsinki), nnet
(commercial, Neural Networks Toolbox, The Mathworks Inc)
Literature:
1. Ch. M. Bishop, Neural networks for Pattern Recognition.
Clarendon Press, Oxford, 1996
2. D. Pissarenko, Neural Networks for Financial Time Series
Prediction: Overview Over Recent Research, pp. 1--186.
http://citeseer.ist.psu.edu/653647.html
3. Robi Polikar, Ensemble base systems in decision making.
IEEE Circuits and Systems Magazine, 2006, 3, 21--45.
Article -- tutorial.
Preprint version downloadable from the authors homepage.
http://users.rowan.edu/~polikar/RESEARCH/PUBLICATIONS/
4. R. Rojas, Neural Networks - A Systematic Introduction. Springer 1996.
s o f t w a r e:
4. Ian Nabney, Netlab: Algorithms for Pattern
Recognition. Springer 2001.
Series: Advances in Pattern Recognition.
NETLAB software downloadable from the authors homepage:
http://www.ncrg.aston.ac.uk/netlab/down.php
5. J. Vesanto, J. Himberg, E. Alhoniemi, J. Parhankangas,
SOM Toolbox for Matlab 5. Som Toolbox Team, Helsinki
University of Technology, Finland, Libella Oy, Espoo 2000, 1--54.
SOM Toolbox software downloadable from:
www.cis.hut.fi/projects/somtoolbox
Other valuable description of different kind of ANN's may be
found a.o. in:
-- R.O. Duda, P.E. Hart, D.G. Stork: Pattern Classification,
2nd Edition, Wiley 2001.
-- Rosaria Silipo, Neural Networks, Chapter 7 from:
M. Bertold, D.J. Hand (Eds.) Intelligent Data Analysis,
Springer Berlin 1999, pp. 217--268.
-- Robi Polikar, Pattern Recognition. Wiley Encyclopedia on
Biomedical Engineering. Copyright 2006 Wiley and Sons, Inc.
Article-survey.
Manuscript downloadable from the authors homepage - contains
some misprints.