Int. J. Environment and Pollution, Vol. 31, Nos. 1/2, 2007 107
Copyright © 2007 Inderscience Enterprises Ltd.
Use of qualitative constraints in modelling of the Lake
Glumsø
Daniel Vladušič*, Boris Kompare
and Ivan Bratko
Faculty of Computer and Information Science,
Trzaska 25, 1000 Ljubljana, Slovenia
Fax: 386 1 476 83 86 E-mail: daniel.vladusic@fri.uni-lj.si
*Corresponding author
Abstract: This paper describes modelling of time behaviour of phytoplankton
and zooplankton in the Danish lake Glumsø with a recently developed
approach to machine learning in numerical domains, called Q
2
learning.
An essential part of this approach is qualitative constraints which were
either handcrafted using knowledge from the Lotka-Volterra predator-prey
model or induced directly from the collected data with a program called QUIN.
The induced models were evaluated by a domain expert. We performed a
comparison between numerical results of the Q
2
learning approach and standard
machine learning algorithms. The results suggest that use of qualitative
constraints leads to more accurate quantitative predictions.
Keywords: qualitative reasoning; machine learning; numerical prediction;
Lake Glumsø.
Reference to this paper should be made as follows: Vladušič, D., Kompare, B.
and Bratko, I. (2007) ‘Use of qualitative constraints in modelling of the Lake
Glumsø’, Int. J. Environment and Pollution, Vol. 31, Nos. 1/2, pp.107–124.
Biographical notes: Daniel Vladušič is a Researcher at the Faculty of
Computer and Information Science, University of Ljubljana, Slovenia. He has
conducted research in machine learning, behavioural cloning and qualitative
modelling. His work is primarily focused on use of qualitative models in
numerical prediction. His work was applied to various modelling tasks from the
field of dynamic systems control, where he modelled active and passive damper
systems, vehicle behaviour and crane control. Using ecological modelling, he
conducted research in the prediction of the river Savinja water level and in the
Lake Glumsø domain.
Boris Kompare is an Associate Professor at the Faculty of Civil and Geodetic
Engineering, University of Ljubljana, where he chairs the Institute of Sanitary
Engineering. His main field of work is sanitary and ecological engineering.
He lectures subjects as Fundamentals of Water Treatment, Water Supply,
Drinking Water Treatment, Water Protection, Ecological Modelling, etc.
His present research topics are the use of machine learning tools in ecological
modelling, improvements of the mercury cycling model, fate of xenobiotic
chemicals in urban water cycle, development of robust and economically
attractive small waste water treatment plants, development of revised hydraulic
friction loss formula, etc.
Ivan Bratko is Professor of Computer Science at the Faculty of Computer
and Information Science, University of Ljubljana, Slovenia. He heads the
AI laboratory at the University. He has conducted research in machine learning,