Evolution of Deep Neural Network Architecture Using Particle Swarm Optimization to Improve the Performance in Determining the Friction Angle of Soil

Table 9

Summary of the 300 simulations using RMSE criteria.

Model	Data set	Average (^o)	Min (^o)	Max (^o)	SD
10-HD-R^2	Training	1.566	1.260	1.848	0.1088
	Validation	2.274	1.782	3.498	0.3395
	Testing	1.77	1.406	2.592	0.1835
10-HD-MAE	Training	2.059	1.421	3.128	0.3711
	Validation	2.668	1.876	5.095	0.473
	Testing	2.255	1.578	3.310	0.2988
8-HD-RMSE	Training	1.722	1.465	3.598	0.2129
	Validation	2.22	1.676	4.867	0.3861
	Testing	2.019	1.613	3.882	0.3264