Mango Grading System Based on Optimized Convolutional Neural Network

Table 3

Different batch size and epoch parameters on training results.

Experiment number	Batch size	Iterations	Epoch	Accuracy (%)	Top1 error	Loss	Running time (s)
1	12	560	80	88.2	0.118	0.44	56
2		700	100	70.6	0.294	0.79	63
3		360	120	84.4	0.124	0.58	76
4	16	480	80	80.0	0.210	1.377	56
5		600	100	86.7	0.113	0.443	70
6		720	120	95.8	0.059	0.441	73
7	20	400	80	76.5	0.235	1.416	68
8		500	100	86.7	0.113	0.444	74
9		600	120	80.2	0.220	1.416	84