Skip to main content

Advertisement

Table 7 Accuracy of classifiers and the number of genes used in classifiers (in parenthesis) for the independent test set for multi-class expression datasets

From: TSG: a new algorithm for binary and multi-class cancer classification and informative genes selection

Method Leuk1 Lung1 Leuk2 SRBCT Breast Lung2 DLBCL Leuk3 Cancers GCM Aver
TSG 97.06
(6)
81.25
(20)
100
(44)
100
(13)
86.67
(63)
95.52
(60)
93.33
(16)
91.07
(95)
79.73
(81)
67.39
(112)
89.20
HC-TSP* 97.06
(4)
71.88
(4)
80
(4)
95
(6)
66.67
(8)
83.58
(8)
83.33
(10)
77.68
(12)
74.32
(20)
52.17
(26)
78.17
HC-k-TSP* 97.06
(36)
78.13
(20)
100
(24)
100
(30)
66.67
(24)
94.03
(28)
83.33
(46)
82.14
(64)
82.43
(128)
67.39
(134)
85.12
DT* 85.29
(2)
78.13
(4)
80
(2)
75
(3)
73.33
(4)
88.06
(5)
86.67
(5)
75.89
(16)
68.92
(10)
52.17
(18)
76.35
NB* 85.29 81.25 100 60 66.67 88.06 86.67 32.14 79.73 52.17 73.2
k-NN* 67.65 75 86.67 30 63.33 88.06 93.33 75.89 64.86 34.78 67.96
PAM* 97.06
(44)
78.13
(13)
93.33
(62)
95
(285)
93.33
(4822)
100
(614)
90
(3949)
93.75
(3338)
87.84
(2008)
56.52
(1253)
88.5
1-vs-1-SVM* 79.41 87.5 100 100 83.33 97.01 100 84.82 83.78 65.22 88.11
  1. *Results reported in Tan et al. [3]Results obtained with our method (TSG) : NB, k-NN, 1-vs-1-SVM used entire set of genes in classification. That is, 7129, 7129, 12582, 2308, 9216, 12600, 4026, 12558, 12533, 16063 for the ten data sets respectively.