The graphic representations of ˆ ( | ) 1 y t m for time errors detection for manufacturing dental pieces manufacturing
-y1(t) - is shown in “Fig. 4” and “Fig. 5” for a feedforward network structure. In “Fig. 4”,
the x-axis shows the number of total samples and the y-axis represents the normalized output variable range:
which are the normalized time errors for manufacturing. In “Fig. 5”,
the y-axis represents the output unnormalized. For the technique of early stopping,
the estimation, validation and test data sets include 23, 6 and 6 samples, respectively.
For the technique of Bayesian regularization, the estimation data set includes 35 samples.