Skip to main content

Table 1 Comparison of classification accuracy on the same platform using top ranked gene isoforms.

From: Evaluation of data discretization methods to derive platform independent isoform expression signatures for multi-class tumor subtyping

Feature selection

CV (#)

SVM-RFE (#)

RF_based_FS (#)

Classifier

FC

Equal-W

Equal-F

k-means

FC

Equal-W

Equal-F

k-means

FC

Equal-W

Equal-F

k-means

SVM

97.6 (1000)

97.6 (600)

96.4 (500)

96.4 (600)

97.6 (1000)

95.2 (600)

97.6 (1000)

97.6 (1000)

98.8 (150)

96.4 (70)

96.4 (200)

98.8 (150)

RF

94.1 (900)

95.2 (1000)

92.9 (800)

91.7 (600)

96.4 (900)

96.4 (1000)

95.2 (400)

96.4 (1000)

95.2 (300)

94.1 (150)

92.9 (500)

92.9 (400)

NB

94.1 (700)

94.1 (900)

92.9 (600)

92.9 (700)

83.5 (1000)

83.5 (1000)

88.2 (80)

83.5 (1000)

95.2 (300)

95.2 (200)

95.2 (600)

96.4 (400)

PAM

89.4 (900)

92.9 (1000)

91.7 (1000)

91.7 (1000)

87.0 (900)

83.5 (900)

87.0 (900)

87.0 (900)

92.9 (600)

94.1 (200)

94.1 (800)

92.9 (400)

  1. # Number of variables in the classification model
  2. Comparison of classification methods both trained (257 samples) and tested (85 samples) on exon-array data. The best accuracy (percentage of samples correctly predicted) achieved by each combination of the four classifiers and three feature selection schemes are presented, with number of features used in the best fitted model is shown in parenthesis. The models were built by stepwise addition of feature variables into the model by considering the top 1,000 ranked feature variables. Best accuracy, achieved with the least number of features, is marked in bold for each classification method.