EARN: an ensemble machine learning algorithm to predict driver genes in metastatic breast cancer

BMC Medical Genomics

Table 5 Validation of four learning methods by some evaluation metrics. (a) MBCA, (b) BRCA

Method name	F1 score	False Positive Rate	Maximum Precision	Average-Precision	Recall	ROC-AUC^a
(a) MBCA
EARN	0.7961	0	1	0.8266	0.6701	0.9924
	SD^b: 0.0264	SD: 0.0	SD: 0.0	SD: 0.0162	SD: 0.0338	SD: 0.0008
RF	0.756	0.0008	0.9069	0.7873	0.6603	0.9418
ANN	0.799	0	1	0.8074	0.6733	0.968
NLSVM	0.3972	0.0154	0.3092	0.5852	0.5885	0.977
(b) BRCA
EARN	0.9313	0	1	0.9585	0.8749	0.9979
	SD: 0.0117	SD: 0.0	SD: 0.0	SD: 0.0079	SD: 0.0193	SD: 0.0005
RF	0.8864	0.0019	0.9061	0.9171	0.8774	0.9719
ANN	0.8996	0	1	0.9417	0.8225	0.9873
NLSVM	0.5441	0.0279	0.446	0.859	0.8422	0.9926

ISSN: 1755-8794