Predikce technologických parametrů s použitím dataminingových metod

Abstract

The reason for this thesis is the prediction of the final product‘s technological parameters, based on contactless measurements during the production process. For the purpose of prediction is utilized instantly learning and adapting neural network, which is converted into programming code fed into the software application. This thesis describes various sections of production and explains contactless measurement providing inputs for subsequent prediction. Description of neural network’s learning procedure is based on gained measured inputs and outputs.