Automatic discovery of 100-miRNA signature for cancer classification using ensemble feature selection

BMC Bioinformatics

Table 7 TNBC classification from the other molecular subtypes in the TCGA dataset, using 1046 features and 100 signature

	TNBC-100	TNBC-1046	Other-100	Other-1046	Global-100	Global-1046
#Samples	135	135	629	629	764	764
Gradient Boosting	0.9111	0.8963	0.9857	0.9857	0.9725	0.9699
Random Forest	0.8889	0.8815	0.9905	0.9905	0.9725	0.9712
Logistic Regression	0.8963	0.9630	0.9793	0.9587	0.9647	0.9593
Passive Aggressive	0.8815	0.9630	0.9714	0.9523	0.9556	0.9540
SGD	0.8000	0.8222	0.9809	0.9841	0.9490	0.9555
SVC	0.8444	0.8963	0.9666	0.9825	0.9451	0.9673
Ridge	0.8000	0.7259	0.9825	0.9237	0.9503	0.8888
Bagging	0.8444	0.8963	0.9793	0.9825	0.9555	0.9673
Average	0.8583	0.8806	0.9795	0.9700	0.9582	0.9542

ISSN: 1471-2105