Multiobjective Genetic Algorithm and Convolutional Neural Network Based COVID-19 Identification in Chest X-Ray Images

Table 2

Testing (i.e., validation) analysis of the COVID-19 identification models by utilizing confusion matrix-based performance metrics when training to testing ratio is 60 : 40.

Model	Accuracy	F-measure	Sensitivity	Specificity	AUC
SVM	0.857627	0.864407	0.863481	0.858586	0.861017
ANFIS	0.869492	0.876271	0.875427	0.870537	0.872881
CNN	0.881356	0.888136	0.887372	0.882155	0.884746
AlexNet	0.893522	0.932324	0.899317	0.893939	0.895661
ResNet-34	0.905085	0.911864	0.911263	0.905724	0.908475
GoogLeNet	0.916949	0.923729	0.923208	0.917508	0.920339
VGG-16	0.928814	0.935593	0.935154	0.929293	0.932203
ResNet-50	0.940678	0.947458	0.947099	0.941077	0.944068
Xception	0.952542	0.959322	0.959044	0.952862	0.955932
DenseNet201	0.964407	0.971186	0.970949	0.964646	0.967797
Proposed model	0.976271	0.983051	0.982935	0.976431	0.979661