Modelling and Prediction of Photovoltaic Power Output Using Artificial Neural Networks

Table 3

The GRNN comparison for different spreads.

Spread	MSE train	MSE test	Regression
0.01	5.16 × 10^−8	0.00072	0.98948
0.02	4.15 × 10^−6	0.00054	0.99238
0.03	2.23 × 10^−5	0.0004	0.99476
0.04	6.80 × 10^−5	0.00039	0.99559
0.05	0.00015	0.00049	0.99505
0.06	0.00029	0.00068	0.99362
0.07	0.00048	0.00094	0.99151
0.08	0.00072	0.00126	0.98882
0.09	0.001	0.00162	0.98575
0.1	0.00131	0.00201	0.98243