TABLE 5.

Comparative analysis of classification quality (accuracy) for classifiers trained without transformation (in initial feature space) and with transformation (in transformed space from the use of a neutral network) on mutational data.

	Accuracy	Kappa	Sensitivity	Specificity	Auroc	Mcnemar’s test p-value
LR	0.8151	0.0694	0.8286	0.6984	0.5635	<0.001
LDA	0.8214	0.0742	0.8286	0.8065	0.5675	<0.001
ANN	0.8559	0.0498	0.9143	0.7855	0.5499	<0.001
SVM	0.7597	0.1134	0.4286	0.7968	0.7513	<0.001