@article{Арсірій_2013, title={Developing of hydro-aerodynamic systems’ elements based on intellectual visualization}, volume={3}, url={https://journals.uran.ua/eejet/article/view/14826}, DOI={10.15587/1729-4061.2013.14826}, abstractNote={<em>For identifying and studying the hydro-aerodynamic processes influence the hydraulic resistance in the elements of a complex shape complex modeling tools are developed. These tools allow receive the visual and intellectual models. For receiving of hydrodynamic intellectual models of hydro-aerodynamic structural primitives in the implementations array of the hydro-aerodynamic flows in major and minor hydro-aerodynamic systems’ elements analysis and search of regularities for a more compact description of the data were carried out. The analysis was carried out in four phases: a priori alphabet classes preparation based on the clustering procedure, a priori attributes dictionary preparation, the classification algorithm implementation; clarify the description of all the classes in the signs language  considering the classification quality index. For determining the prototype classes it was proposed to carry out the clustering of the visual data based on the method of intelligent visualization (MIV) using Kohonen self-organizing maps (SOM) in the form of coherent regions (coherent region) (SOM-CR).MIV SOM-CR is implemented in the form of two procedures: the classical self-organization and the modified calibration.At the stage of classification for separating surfaces building the multilayer perceptron (multilayer perceptron – MLP) is used. It was trained on the basis of back-propagation (back-propagation learning – BPL)</em>}, number={8(63)}, journal={Eastern-European Journal of Enterprise Technologies}, author={Арсірій, Олена Олександрівна}, year={2013}, month={Jun.}, pages={4–8} }