The decision support system for the classification of allergenic pollen types based on fuzzy expert data of pollen features on the microscope images

The article is devoted to the development of the decision support system for the classification of allergic pollen types based on fuzzy expert data of their features on the microscope images and the improvement of the identification method of this system. The genetic algorithm of the classifier training has been improved taking into account the spatial-temporal features of the distribution of pollen in the air, namely: adaptive fitness function for evaluating solutions, tournament selection with elitism and initialization of the initial population, allowing to improve the accuracy of classification. The example of the proposed system and method application based on real expert data of Vinnytsia monitoring station which is the part of European Aeroallergen Network is shown.