The goal of the fuzzy diatoms models is to use the advancement of the fuzzy theory and using the predictive modelling to predict the outcome of given condition of diatoms. Fuzzy modelling depends from many factors, mainly by the shape of the membership functions, similarity metrics and above all the operations between the fuzzy sets – the fuzzy operators. In this paper we introduce new fuzzy operator (fuzzy geometric operator) that considers another property of the ecological datasets – the ration measurements. In this way, the ecological dataset includes important property that we expect to increase the classification accuracy of the models. The use of simpler membership functions like trapezoidal or triangular could be more appropriate with this type of operator compared with the average geometric operator used previously. The evaluation results compared between the operators have confirmed these findings. Based on these results, one model for each water quality category class is presented and discussed.
fuzzy membership functions, geometric operator, diatoms, water quality models.