[Retracted] The Application of Pattern Recognition System in Design Field Based on Aesthetic Principles

Table 1

A 12-column convolutional neural network trained on the MNIST dataset.

Convolutional network	L1-C	L2-P	L3-C	L4-P	Error rate (%)
CNN1	6(5)	2	12(5)	2	0.98
CNN2	6(9)	2	12(5)	2	1.00
CNN3	6(13)	2	12(5)	2	1.46
CNN4	6(5)	2	24(5)	2	0.91
CNN5	6(9)	2	24(5)	2	0.85
CNN6	6(13)	2	24(5)	2	1.06
CNN7	12(5)	2	24(5)	2	0.78
CNN8	12(9)	2	24(5)	2	0.66
CNN9	12(13)	2	24(5)	2	0.92
CNN10	16(5)	2	24(5)	2	0.65
CNN11	16(9)	2	24(5)	2	0.69
CNN12	16(13)	2	24(5)	2	0.76