ASSESSING THE QUALITY OF NEURAL MODELS USING A MODEL OF FLOW CHARACTERISTICS OF FABRICS AS AN EXAMPLE

Abstract

In this article, a method for assessing the quality of now-widespread neuron models is presented. Attention has been paid to the significance of analysing individual input quantities in the model constructed. Then, the parameters which are quantitative measures of the neural model’s quality have to be specified. The need for structural verification of the network is stressed, which is the basis for stating that the neural model obtained has been well matched with the investigation results. The assessment of the neural model’s quality has been shown, using a model of the flow characteristics of fabrics as an example.