Systematische Optimierung von Modellen für Automatisierungstechnik und IT
Das BMBF-geförderte Projekt SOMA wurde im Zeitraum 2009-2013 durchgeführt. Projektpartner waren die Universität Leiden, die Ruhr-Universität Bochum, die Nurogames GmbH, Köln und die divis GmbH, Dortmund.
Die systematische Optimierung von Modellen für komplexe Anwendungen in Informations- und Automatisierungstechnik, hier mit dem Ziel der Prognose von Zielgrößen oder der optimalen Steuerung von Anlagen oder Prozessen, ist Gegenstand dieses Projektes. Sie stellt nach wie vor eine große Herausforderung für den in der Praxis tätigen Informatiker oder Ingenieur dar. In vielen Fällen handelt es sich nicht allein um ein Problem der optimalen Modellparametrierung, sondern auch um Fragen der intelligenten Datenvorverarbeitung und Datenselektion.
Unter dem Dach von SOMA wurden verschiedene Unterprojekte bearbeitet: Tuned Data Mining (TDMR), Gestenerkennung, Reinforcement Learning für strategische Spiele, Intelligente Methoden der Merkmalsgewinnung, wie z.B. Slow Feature Analysis (SFA). 
 
Themengebiete: Angewandte Informatik, Modellierung, Simulation, Lernende Systeme, Computational Intelligence (evolutionäre Algorithmen, neuronale Netze), Data Mining.
 

Publicationen SOMA

2018

Konen, Wolfgang; Koch, Patrick

The TDMR 2.0 Package: Tuned Data Mining in R Technical Report

Research Center CIOP (Computational Intelligence, Optimization and Data Mining) Cologne University of Applied Science, Faculty of Computer Science and Engineering Science, Nr. 02/2018, 2018, (Last update: April 2018 (original version: 2012)).

Links | BibTeX

Konen, Wolfgang; Koch, Patrick

The TDMR 2.0 Tutorial: Examples for Tuned Data Mining in R Technical Report

Research Center CIOP (Computational Intelligence, Optimization and Data Mining) Cologne University of Applied Science, Faculty of Computer Science and Engineering Science, Nr. 03/2018, 2018, (Last update: April 2018 (original version: 2012)).

Links | BibTeX

2015

Stork, Jörg; Ramos, Ricardo; Koch, Patrick; Konen, Wolfgang

SVM ensembles are better when different kernel types are combined Book Section

In: Lausen, Berthold; Krolak-Schwerdt, Sabine (Hrsg.): Data Science, Learning by Latent Structures, and Knowledge Discovery, S. 191–201, Springer, 2015.

Links | BibTeX

2013

Koch, Patrick; Konen, Wolfgang

Subsampling strategies in SVM ensembles Proceedings Article

In: Hoffmann, Frank; Hüllermeier, Eyke (Hrsg.): Proceedings 23. Workshop Computational Intelligence, S. 119–134, Universitätsverlag Karlsruhe, 2013.

Links | BibTeX

Stork, Jörg; Ramos, Ricardo; Koch, Patrick; Konen, Wolfgang

SVM ensembles are better when different kernel types are combined Proceedings Article

In: Lausen, Berthold (Hrsg.): European Conference on Data Analysis (ECDA13), GfKl, 2013.

Links | BibTeX

2012

Konen, Wolfgang; Koch, Patrick

The TDMR Package: Tuned Data Mining in R Technical Report

Research Center CIOP (Computational Intelligence, Optimization and Data Mining) Cologne University of Applied Science, Faculty of Computer Science and Engineering Science, Nr. 02/2012, 2012, (Last update: June 2017).

Links | BibTeX

Konen, Wolfgang; Koch, Patrick

The TDMR Tutorial: Examples for Tuned Data Mining in R Technical Report

Research Center CIOP (Computational Intelligence, Optimization and Data Mining) Cologne University of Applied Science, Faculty of Computer Science and Engineering Science, Nr. 03/2012, 2012, (Last update: May, 2016).

Links | BibTeX

Koch, Patrick; Konen, Wolfgang

Efficient sampling and handling of variance in tuning data mining models Proceedings Article

In: Coello Coello, Carlos; Cutello, Vincenzo; others, (Hrsg.): PPSN'2012: 12th International Conference on Parallel Problem Solving From Nature, Taormina, S. 195–205, Springer, Heidelberg, 2012.

Links | BibTeX

Thill, Markus; Koch, Patrick; Konen, Wolfgang

Reinforcement learning with n-tuples on the game Connect-4 Proceedings Article

In: Coello Coello, Carlos; Cutello, Vincenzo; others, (Hrsg.): PPSN'2012: 12th International Conference on Parallel Problem Solving From Nature, Taormina, S. 184–194, Springer, Heidelberg, 2012.

Links | BibTeX

Koch, Patrick; Bischl, Bernd; Flasch, Oliver; Bartz-Beielstein, Thomas; Weihs, Claus; Konen, Wolfgang

Tuning and Evolution of Support Vector Kernels Journal Article

In: Evolutionary Intelligence, Bd. 5, S. 153–170, 2012.

Links | BibTeX

2011

Konen, Wolfgang

Der SFA-Algorithmus für Klassifikation Technical Report

Research Center CIOP (Computational Intelligence, Optimization and Data Mining) Cologne University of Applied Science, Faculty of Computer Science and Engineering Science, Nr. 08/11, 2011, ISSN: 2191-365X.

Links | BibTeX

Konen, Wolfgang

SFA classification with few training data: Improvements with parametric bootstrap Technical Report

Research Center CIOP (Computational Intelligence, Optimization and Data Mining) Cologne University of Applied Science, Faculty of Computer Science and Engineering Science, Nr. 09/11, 2011, ISSN: 2191-365X.

Links | BibTeX

Koch, Patrick; Bischl, Bernd; Flasch, Oliver; Bartz-Beielstein, Thomas; Konen, Wolfgang

On the Tuning and Evolution of Support Vector Kernels Technical Report

Research Center CIOP (Computational Intelligence, Optimization and Data Mining) Cologne University of Applied Science, Faculty of Computer Scienceand Engineering Science, Nr. 04/11, 2011, ISSN: 2191-365X.

Links | BibTeX

Konen, Wolfgang; Koch, Patrick; Flasch, Oliver; Bartz-Beielstein, Thomas; Friese, Martina; Naujoks, Boris

Tuned Data Mining: A Benchmark Study on Different Tuners Technical Report

Cologne University of Applied Sciences Nr. 03/11, 2011.

Links | BibTeX

Koch, Patrick; Konen, Wolfgang; Naujoks, Boris; Flasch, Oliver; Friese, Martina; Zaefferer, Martin; Bartz-Beielstein, Thomas

Tuned Data Mining in R Proceedings Article

In: Hoffmann, Frank; Hüllermeier, Eyke (Hrsg.): Proceedings 21. Workshop Computational Intelligence, S. 147–160, Universitätsverlag Karlsruhe, 2011.

Links | BibTeX

Konen, Wolfgang; Koch, Patrick; Flasch, Oliver; Bartz-Beielstein, Thomas; Friese, Martina; Naujoks, Boris

Tuned Data Mining: A Benchmark Study on Different Tuners Proceedings Article

In: Krasnogor, Natalio (Hrsg.): GECCO '11: Proceedings of the 13th Annual Conference on Genetic andEvolutionary Computation, S. 1995–2002, 2011.

BibTeX

Konen, Wolfgang; Koch, Patrick

The slowness principle: SFA can detect different slow components in nonstationary time series Journal Article

In: International Journal of Innovative Computing and Applications (IJICA), Bd. 3, Nr. 1, S. 3–10, 2011.

Links | BibTeX

2010

Koch, Patrick; Konen, Wolfgang; Hein, Kristine

Gesture Recognition on Few Training Data using Slow Feature Analysis and Parametric Bootstrap Proceedings Article

In: 2010 International Joint Conference on Neural Networks, 2010.

Links | BibTeX

Koch, Patrick; Konen, Wolfgang; Flasch, Oliver; Bartz-Beielstein, Thomas

Optimizing Support Vector Machines for Stormwater Prediction Proceedings Article

In: Bartz-Beielstein, Thomas; Chiarandini, M.; Paquete, L.; Preuss, Mike (Hrsg.): Proceedings of Workshop on Experimental Methods for the Assessment of Computational Systems joint to PPSN2010, S. 47–59, TU Dortmund, 2010.

Links | BibTeX

Koch, Patrick; Konen, Wolfgang; Flasch, Oliver; Bartz-Beielstein, Thomas

Optimization of Support Vector Regression Models for Stormwater Prediction Proceedings Article

In: Hoffmann, Frank; Hüllermeier, Eyke (Hrsg.): Proceedings 20. Workshop Computational Intelligence, S. 146–160, Universitätsverlag Karlsruhe, 2010.

BibTeX

Konen, Wolfgang; Koch, Patrick; Flasch, Oliver; Bartz-Beielstein, Thomas

Parameter-Tuned Data Mining: A General Framework Proceedings Article

In: Hoffmann, Frank; Hüllermeier, Eyke (Hrsg.): Proceedings 20. Workshop Computational Intelligence, Universitätsverlag Karlsruhe, 2010.

Links | BibTeX

Konen, Wolfgang; Koch, Patrick

How slow is slow? SFA detects signals that are slower than the driving force Proceedings Article

In: Filipic, Bogdan; Silc, Juri (Hrsg.): Proc. 4th Int. Conf. on Bioinspired Optimization Methods and their Applications, BIOMA, Ljubljana, Slovenia, 2010.

Links | BibTeX

Hein, Kristine

Gestenerkennung mit Slow Feature Analysis (SFA) - Klassifizierung von beschleunigungsbasierten 3D-Gesten des Wii-Controllers Masters Thesis

TH Köln -- University of Applied Sciences, 2010, (Master thesis, 3rd prize in Opitz award 2011).

Links | BibTeX

2009

Konen, Wolfgang

How slow is slow? SFA detects signals that are slower than the driving force Technical Report

Cologne University of Applied Sciences Nr. 05/09, 2009, (e-print published at http://arxiv.org/abs/0911.4397).

Links | BibTeX

Konen, Wolfgang

On the numeric stability of the SFA implementation sfa-tk Technical Report

Cologne University of Applied Sciences Nr. 05/10, 2009, (e-print published at http://arxiv.org/abs/0912.1064).

Links | BibTeX