Skip to main content

Table 5 The diagnostic results with independent training and test sets for LGG data

From: Diagnostic biases in translational bioinformatics

Algorithm Accuracy ± std (%) Sensitivity ± std (%) Specificity ± std (%) NPR ± std (%) PPR ± std (%)
   GliomaRNASeq    
DCA-SVM 99.52 ± 00.58 99.64 ± 00.53 97.00 ± 08.08 91.18 ± 11.05 99.87 ± 00.36
SVM-linear 95.87 ± 00.84 98.77 ± 00.00 12.10 ± 11.46 NaN 97.02 ± 00.85
SVM-rbf 96.68 ± 00.78 100.0 ± 00.00 00.00 ± 00.00 NaN 96.68 ± 00.78
SVM-quad 96.53 ± 00.75 99.60 ± 00.40 07.40 ± 08.67 NaN 96.91 ± 00.79
SVM-mlp 56.28 ± 05.61 56.73 ± 05.95 43.77 ± 18.61 03.34 ± 01.53 96.70 ± 01.31
   GliomaMiRNASeq    
DCA-SVM 99.63 ± 00.52 99.73 ± 00.39 97.52 ± 08.21 93.13 ± 09.40 99.89 ± 00.39
SVM-linear 93.78 ± 01.27 96.68 ± 01.58 10.93 ± 10.04 10.89 ± 11.08 96.89 ± 00.84
SVM-rbf 96.63 ± 00.81 100.0 ± 00.00 00.00 ± 00.00 NaN 96.63 ± 00.81
SVM-quad 95.65 ± 00.97 98.76 ± 00.87 06.14 ± 07.39 NaN 96.80 ± 00.79
SVM-mlp 56.62 ± 06.31 58.16 ± 06.69 42.51 ± 18.98 03.45 ± 01.74 96.68 ± 01.30