Conjugate Cellular Automata and Neural Network Approach: Failure Load Prediction of Masonry Panels

Table 5

Comparison between NN, FEA, and YL accuracy of prediction.

Panel	YL error (%)	FEA error (%)	NN error (%)	Panel	YL error (%)	FEA error (%)	NN error (%)
1120	20.73	11.47	0.69	1150	3.27	17.48	0.26
1135	51.85	5.38	42.92	1153	35.22	5.07	10.86
1116	12.41	15.57	17.08	1237	26.77	4.62	1.44
1126	38.75	4.22	2.36	1148	39.81	2.72	1.63
1190	32.19	3.91	1.26	1211	58.20	16.41	0.78
1108	28.08	6.35	3.00	1246	63.47	27.88	2.57
1109	18.75	13.44	4.03	1170	15.81	12.44	3.93
1187	20.71	16.79	89.39	1178	68.46	14.04	21.40
1094	40.00	13.33	17.93	1224	68.96	14.38	64.00
1110	0.80	18.40	15.09	1149	16.67	14.70	11.15
1095	4.48	32.76	14.37	1162	48.14	4.65	4.74
1171	9.55	24.09	63.32	1231	50.26	0.51	60.82
1121	18.62	19.14	6.48	ART01	–	3.64	2.50
1123	21.59	30.91	57.11	SB01	13	12.00	6.01
1117	15.81	29.77	6.27	SB05	17	9.00	17.93
1203	39.41	12.65	24.12	SB06	16	10.00	21.06
1107	31.40	53.80	10.59	Panels with an opening
1111	52.09	72.79	41.22	Panel 1	–	12.67	17.95
1201	10.95	51.90	44.98	Panel 2	–	1.67	25.29
1096	41.07	32.50	20.47	Panel 3	–	38.00	6.26
1097	16.67	49.44	70.38	ART02	–	13.09	9.57
1157	3.57	47.86	31.44	ART03	–	0.00	12.79
1172	50.00	13.64	2.53	ART04	–	11.60	2.29
1192	49.36	25.53	0.33	ART06	–	7.60	2.53
1261	81.73	57.27	1.59	SB02	8	24.00	2.55
1169	51.58	22.74	5.89	SB03	6	31.00	7.74
1173	53.88	18.63	34.07	SB04	17	18.00	18.26
1244	41.00	7.88	1.40	SB07	27	6.00	0.25