A methodology to assess and improve the physics consistency of an artificial neural network regression model for engineering applications

Advanced Modeling and Simulation in Engineering Sciences

Table 2 Regions of acceptable approximation for 2nd degree Taylor polynomial

Development points	\({\Delta xc}_{a1}\)	\({\Delta xc}_{a2}\)	\({\Delta xc}_{a3}\)	\({\Delta xc}_{a4}\)	\({\Delta xc}_{a5}\)
Point A	0.037500	0.037500	0.046875	0.046875	0.084375
Point B	0.032812	0.032812	0.032812	0.065625	0.109375
Point C	0.028125	0.028125	0.036562	0.079062	0.126562
Mean	0.328123	0.328123	0.038749	0.063854	0.106770
Std deviation	0.003827	0.003827	0.005945	0.013199	0.017320
\(\mu -\sigma (\gamma )\)	0.324296	0.324296	0.032804	0.050655	0.08945