Applications of artificial neural networks predicting macroinvertebrates in freshwaters

Abstract

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.

Modelling Tools to Analyze and Assess the Ecological Impact of Hydropower Dams

“…Apart from increased data availability, there has been substantial progress made in ecological modelling with regard to low quality or quantity data. When the amount of data is limited, datasets could be split into a training and validation set [133] in order to test the robustness of the models. Moreover, independent experts could check model performance or the prediction of the model could be evaluated if the model output lies within the reliable ecological limits [120].…”

A comparison of artificial neural networks and random forests to predict native fish species richness in Mediterranean rivers

“…To minimize the error, the algorithm employs the values of the error calculated in the previous iteration and then updates the weights. A detailed description of ANN is reported in Olden et al (2008) and Goethals et al (2007). This study applied a MLP to predict native fish species richness.…”

“…LM has been applied successfully in ecology (Gutiérrez-Estrada and Bilton, 2010) and a description of the algorithm can be found in Singh et al (2009). To test and validate the models, we used the k-fold crossvalidation procedure and tested different k values (ranging from 3 to 10, Goethals et al, 2007;Dormann, 2011). The best k value was then identified by the comparison of the performance of the different ANN obtained in the cross-validation procedure.…”

“…The best k value was then identified by the comparison of the performance of the different ANN obtained in the cross-validation procedure. This approach is frequently used when the number of data available is not sufficient to divide the data into training and validation datasets (Goethals et al, 2007;Olden et al, 2008) The dataset is randomly split in k subsets: k-1 folds are used for training and the remaining used for validation (Hastie et al, 2009) This procedure is repeated k times and results in k models. The performance of the different k models was averaged as final criterion for model evaluation.…”

Utilização de Algoritmo Genético e Redes Neurais Artificiais para Classificação de Blastocistos Bovinos

“…Não mensurando variáveis objetivas, a determinação da classificação do embrião torna-se subjetiva e com pouca repetibilidade [2], no entanto, a aplicação de Redes Neurais Artificiais (RNA) como forma de reconhecimento de padrões [4,5] pode evitar esta subjetividade.…”