Research Article

A Systematic Evaluation of Feature Selection and Classification Algorithms Using Simulated and Real miRNA Sequencing Data

Table 4

Summary of three classification methods using real data.

DatasetsFSLogistic regressionRFSVM
PPVNPVAUCPPVNPVAUCPPVNPVAUC

BRCAbaySeq0.530.530.531.000.990.990.950.960.96
DESeq0.700.720.701.000.991.001.000.940.97
edgeR0.540.550.551.000.990.990.540.550.55
Lasso0.970.980.981.000.990.990.970.980.98
Rank sum0.550.550.551.000.990.990.550.550.55
PSODT0.850.860.860.990.980.980.850.860.86

HNSCbaySeq0.350.380.370.540.560.550.630.520.58
DESeq0.520.570.550.530.520.520.910.470.69
edgeR0.320.350.330.540.540.540.320.350.33
Lasso0.520.760.640.550.550.550.520.760.64
Rank sum0.350.310.330.540.540.540.350.310.33
PSODT0.430.440.430.550.540.540.430.440.43

KICHbaySeq0.360.380.370.650.660.660.680.700.69
DESeq0.370.390.380.660.650.660.680.840.76
edgeR0.400.380.390.660.650.660.400.380.39
Lasso0.640.820.730.650.660.650.640.820.73
Rank sum0.390.380.390.660.660.660.390.380.39
PSODT0.370.380.370.660.650.660.370.380.37

LUADbaySeq0.400.470.430.460.450.460.450.690.57
DESeq0.300.780.540.460.410.440.950.360.65
edgeR0.440.470.460.470.450.460.440.470.46
Lasso0.470.740.610.470.450.460.470.740.61
Rank sum0.300.360.330.470.450.460.300.360.33
PSODT0.360.500.430.470.450.460.360.500.43

STADbaySeq0.420.560.490.440.450.440.440.630.54
DESeq0.140.850.490.410.380.400.910.250.58
edgeR0.370.420.400.490.460.470.370.420.40
Lasso0.430.770.600.460.460.460.430.770.60
Rank sum0.400.480.440.440.460.450.400.480.44
PSODT0.360.440.440.440.460.450.360.440.40

THCAbaySeq0.490.630.560.560.570.570.770.500.63
DESeq0.490.850.670.540.580.560.540.820.68
edgeR0.530.590.560.560.600.580.530.590.56
Lasso0.540.880.710.560.590.570.540.880.71
Rank sum0.440.440.440.570.580.580.440.440.44
PSODT0.480.560.520.560.560.560.480.560.52