Table 1

Performance of classification algorithms for the leukemia data based on 20 repetitions of 10-fold CV

Algorithm

Accuracy

Sensitivity*

Specificity^†

PPV^‡

NPV^§

CERP

98.6 (<.001)

96.0 (<.001)

100.0 (.000)

100.0 (.000)

97.9 (<.001)

RF

97.9 (.008)

95.0 (.022)

99.5 (.009)

99.0 (.018)

97.4 (.011)

AdaBoost

96.0 (.005)

95.6 (.012)

96.3 (.009)

93.2 (.016)

97.6 (.006)

SVM

97.2 (.012)

92.0 (.034)

100.0 (.000)

100.0 (.000)

95.9 (.017)

DLDA

97.5 (.007)

96.0 (<.001)

98.3 (.011)

96.8 (.021)

97.9 (<.001)

SC

96.0 (.004)

96.0 (<.001)

96.0 (.007)

92.7 (.011)

97.8 (<.001)

CART

81.7 (.035)

76.2 (.046)

84.6 (.053)

72.4 (.067)

87.0 (.021)

CRUISE

86.8 (.021)

79.8 (.040)

90.5 (.029)

82.0 (.044)

89.4 (.018)

QUEST

86.9 (.020)

79.4 (.042)

91.0 (.032)

82.7 (.048)

89.3 (.018)

SD is given in parentheses. *AML considered positive. ^†ALL considered negative. ^‡Positive predictive value. ^§Negative predictive value.

Moon et al. Genome Biology 2006 7:R121   doi:10.1186/gb-2006-7-12-r121