Keywords
fuzzy logic, grayling, species distribution modelling, model complexity, data driven
Start Date
1-7-2008 12:00 AM
Abstract
In recent years, fuzzy logic has been acknowledged as a suitable approach for species distribution modelling due to its transparency and its ability to incorporate the ecological gradient theory. Specifically, the overlapping class boundaries of a fuzzy model are similar to the transitions between different environmental conditions. However, the need for ecological expert knowledge is an important constraint when applying fuzzy species distribution models. Recent research has shown that data-driven fuzzy models may solve this ‘knowledge acquisition bottleneck’ and this paper is a further contribution. The aim was to reduce the complexity of a data-driven fuzzy habitat suitability model for European grayling (Thymallus thymallus) in the Aare River (Thun, Switzerland). Therefore, we applied an entropy-based fuzzy set selection algorithm, which allowed minimisation of the number of fuzzy sets needed for data-driven fuzzy model development. Comparison of the presented model with a previously developed model revealed that the entropy-based algorithm reduced model complexity substantially without a significant decrease in predictive accuracy. The results of this study could minimise monitoring costs and efforts, and enhance communication between water managers and stakeholders due to increased model transparency.
Entropy-based fuzzy set optimisation for reducing ecological model complexity
In recent years, fuzzy logic has been acknowledged as a suitable approach for species distribution modelling due to its transparency and its ability to incorporate the ecological gradient theory. Specifically, the overlapping class boundaries of a fuzzy model are similar to the transitions between different environmental conditions. However, the need for ecological expert knowledge is an important constraint when applying fuzzy species distribution models. Recent research has shown that data-driven fuzzy models may solve this ‘knowledge acquisition bottleneck’ and this paper is a further contribution. The aim was to reduce the complexity of a data-driven fuzzy habitat suitability model for European grayling (Thymallus thymallus) in the Aare River (Thun, Switzerland). Therefore, we applied an entropy-based fuzzy set selection algorithm, which allowed minimisation of the number of fuzzy sets needed for data-driven fuzzy model development. Comparison of the presented model with a previously developed model revealed that the entropy-based algorithm reduced model complexity substantially without a significant decrease in predictive accuracy. The results of this study could minimise monitoring costs and efforts, and enhance communication between water managers and stakeholders due to increased model transparency.
