From “Fig. 4”, it can be concluded that the feedforward networks are able to simulate and predict the behaviour of
time errors for dental pieces manufacturing –as a consequence of the production process-
and it is capable of modelling more than 89% of the actual measurements.
In “TABLE II” the value of the indexes obtained for the models are shown.
The model thus obtained can be used not only to predict time errors for dental pieces manufacturing
but also to determine the normal operating conditions of dental milling process.