Table 5.

Parameters of Convolutional Neural Network (CNN) regression models and corresponding RMSE values (Val model optimized using simulated validation set, Exp model optimized using the experimental data).

Parameter	Lift-off		Thickness
	Val	Exp	Val	Exp
RMSE (Val)	0.02	0.02	0.11	0.11
RMSE (Exp, 15)	0.12	0.15	0.38	0.40
RMSE (Exp, 22)	0.06	0.08	0.29	0.24
Layer 1	1D convolution, kernel size = 2, $100\times 50$, ReLU
Layer 2	Flatten
Layer 3	Dense, $50\times 50$, ReLU
Layer 4	Dense, $50\times 1$, Linear
Batch size	256
Loss function	mse
Epochs	100
Optimization	Adam