Modelling water quality, bioindication and population dynamics in lotic ecosystems using neural networks

Schleiter, Ingrid M.; Borchardt, Dietrich; Wagner, Rüdiger; Dapper, Thomas; Schmidt, Klaus-Dieter; Schmidt, Hans-Heinrich; Werner, Heinrich

doi:10.1016/s0304-3800(99)00108-8

Cited by 87 publications

(46 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, the variables have to be scaled in such a way as to be commensurate with the limits of the activation functions used in the output layer (Maier and Dandy 2000). Several authors (Chon et al , 2002Gabriels et al 2007;Obach et al 2001;Park et al 2003a, b;Schleiter et al 1999;Schleiter et al 2001;Wagner et al 2000) proportionally normalized the data between zero and one [0 1] in the range of the maximum and minimum values, while Dedecker et al (2004Dedecker et al ( , 2005a and Gabriels et al (2002) used the interval [À1 1]. Moreover the division of the dataset in folds for cross-validation is crucial for a good model training and evaluation.…”

Section: Data Processingmentioning

confidence: 99%

Applications of artificial neural networks predicting macroinvertebrates in freshwaters

et al. 2007

View full text Add to dashboard Cite

To facilitate decision support in freshwater ecosystem protection and restoration management, habitat suitability models can be very valuable. Data driven methods such as artificial neural networks (ANNs) are particularly useful in this context, seen their time-efficient development and relatively high reliability. However, specialized and technical literature on neural network modelling offers a variety of model development criteria to select model architecture, training procedure, etc. This may lead to confusion among ecosystem modellers and managers regarding the optimal training and validation methodology. This paper focuses on the analysis of ANN development and application for predicting macroinvertebrate communities, a species group commonly used in freshwater assessment worldwide. This review reflects on the different aspects regarding model development and application based on a selection of 26 papers reporting the use of ANN models for the prediction of macroinvertebrates. This analysis revealed that the applied model training and validation methodologies can often be improved and moreover crucial steps in the modelling process are often poorly documented. Therefore, suggestions to improve model development, assessment and application in ecological river management are presented. In particular, data pre-processing determines to a high extent the reliability of the induced models and their predictive relevance. This also counts for the validation criteria, that need to be better tuned to the practical simulation requirements. Moreover, the use of sensitivity methods can help to extract knowledge on the habitat preference of species and allow peer-review by ecological experts. The selection of relevant input variables remains a critical challenge as well. Model coupling is a missing crucial step to link human activities, hydrology, physical habitat conditions, water quality and ecosystem status. This last aspect is probably the most valuable aspect to enable decision support in water management based on ANN models.

show abstract

Section: Data Processingmentioning

confidence: 99%

Applications of artificial neural networks predicting macroinvertebrates in freshwaters

et al. 2007

View full text Add to dashboard Cite

show abstract

“…Sensitivity analysis: A number of investigators have used sensitivity analysis to determine the spectrum of input variable contributions in neural networks. For example, the Senso-nets approach includes an additional weight in the network for each input variable representing the variable's sensitivity (Schleiter et al, 1999). Scardi and Harding (1999) added a white noise to each input variable and examined the resulting changes in the mean square error of the output.…”

Section: Input Significance Testingmentioning

confidence: 99%

Square Surface Aerator: Process Modeling and Parameter Optimization

Rao

Kumar²

2007

J ENV INFORM

View full text Add to dashboard Cite

ABSTRACT. The performance of an aeration system generally depends upon the geometric and dynamic parameters. The main purpose of doing experiment in the area of surface aeration system is to interpret the laboratory results for the field application that is to scale-up the results. This requires a geometrical similarity conditions. Finding geometrical optimal conditions of a surface aeration system through experiments involves physical constraints and classically parameters can be optimized by varying one variable at a time and keeping others as constants. In the real experimental process, it is not possible to vary all others geometric parameters simultaneously. In such a case, the model of the system is built through computer simulation, assuming that the model will result in adequate determination of the optimum conditions for the real system. In this paper, two approaches have been used to model the phenomena: i) Multiple regression and ii) Neural network. It has been found that neural network approach is showing better predictability compared to the multiple regression approach. In process of optimization, the pertinent dynamic parameter is divided into a finite number of segments over the entire range of observations. For each segment of the dynamic parameter, the neural network model is optimized for the geometrical parameters spanning over the entire range of observations. Thus each segment of the dynamic parameter has its set of optimal geometrical conditions. Results obtained are having less variation among them and they are very nearer to the experimental optimal conditions. Input parameter significance test of neural network model reveals that, in general the blade width of rotor is a major geometric parameter to enhance the aeration process.

show abstract

“…Senso-nets include an additional weight for each input neuron representing the relevance (sensitivity) of the corresponding input parameter for the neural network model. These weights are adapted during the training process of the network (Schleiter et al 1999). By using the senso-net, the variables of major importance for the predictions can be selected because the higher the weight of an input variable is, the higher its importance on the absence or presence of a particular macroinvertebrate taxon.…”

Section: Sensitivity Analysismentioning

confidence: 99%

“…In this context, Artificial Neural Networks (ANNs) have been recognized by several authors as valuable tools for predicting macroinvertebrates (e.g. Walley & Fontama 1998, Schleiter et al 1999, Wagner et al 2000, Hoang et al 2001, Park et al 2001, Arab et al 2004, Dedecker et al 2004, 2005a,b, 2006. On the other hand, ANN models have been labelled as 'black boxes' (Olden & Jackson 2002) because they do not provide direct insight in the habitat preferences of the taxa.…”

Section: Introductionmentioning

confidence: 99%

Development of artificial neural network models predicting macroinvertebrate taxa in the river Axios (Northern Greece)

Dakou

Goethals

D'Heygere

et al. 2006

Ann. Limnol. - Int. J. Lim.

View full text Add to dashboard Cite

Artificial Neural Network models (ANNs) were used to predict habitat suitability for 12 macroinvertebrate taxa, using environmental input variables. This modelling technique was applied to a dataset of 102 measurement series collected in 31 sampling sites in the Greek river Axios. The database consisted of seven physical-chemical and seven structural variables, as well as abundances of 90 macroinvertebrate taxa. A seasonal variable was included to allow the description of potential temporal changes in the macroinvertebrate communities. The induced models performed well for predicting habitat suitability of the macroinvertebrate taxa. Senso-nets and sensitivity analyses revealed that dissolved oxygen concentration and the substrate composition always played a crucial role in predicting habitat suitability of the macroinvertebrates. Although ANNs are often referred to as black box prediction techniques, it was demonstrated that ANNs combined with sensitivity analyses can provide insight in the relationship between river conditions and the occurrence of macroinvertebrates, and thus deliver new ecological knowledge. Consequently, these models can be useful in decision-making for river restoration and conservation management.

show abstract

Modelling water quality, bioindication and population dynamics in lotic ecosystems using neural networks

Cited by 87 publications

References 30 publications

Applications of artificial neural networks predicting macroinvertebrates in freshwaters

Applications of artificial neural networks predicting macroinvertebrates in freshwaters

Square Surface Aerator: Process Modeling and Parameter Optimization

Development of artificial neural network models predicting macroinvertebrate taxa in the river Axios (Northern Greece)

Contact Info

Product

Resources

About