Multimodal hybrid convolutional neural network based brain tumor grade classification

BMC Bioinformatics

Table 7 All models’ training performance with a batch size of 16

Model	No of iterations (epoch)	Loss		Rate of error	Testing accuracy (%)
Model	No of iterations (epoch)	Training	Testing	Rate of error	Testing accuracy (%)
Visual Geometry Group “16”	5	0.078	0.389	0.09	93.11
	10	0.071	0.332	0.09	93.23
	15	0.068	0.274	0.08	94.11
	20	0.069	0.201	0.07	94.39
	25	0.056	0.197	0.06	94.87
DenseNet101	10	0.038	0.511	0.07	94.12
	15	0.031	0.412	0.06	94.78
	20	0.028	0.388	0.05	95.05
	25	0.025	0.361	0.04	95.15
Visual Geometry Group “19”	5	0.101	0.119	0.071	94.98
	10	0.081	0.104	0.048	95.82
	15	0.069	0.092	0.041	96.13
	20	0.033	0.081	0.036	96.87
	25	0.027	0.081	0.032	96.77
DenseNet101	10	0.059	0.079	0.05	96.79
	15	0.051	0.071	0.05	97.08
	20	0.048	0.062	0.03	98.12
	25	0.041	0.058	0.03	98.34

ISSN: 1471-2105