Skip to main content

Table 11 Average Accuracy of Each Methods with Results of Wilcoxon Signed-Rank Test

From: Genetic algorithm-based feature selection with manifold learning for cancer classification using microarray data

Dataset

Proposed feature selection framework

CER-ABC (SVM)

CER-ABC (NN)

MBEGA

Iso-GA (SVM)

Iso-GA (NN)

MDS-GA (SVM)

MDS-GA (NN)

GA (SVM)

GA (NN)

Breast

0.821 Ref1 (0.069)

0.735 Ref2 (0.080)

0.726 1 ** (0.096)

0.769 2 ns (0.125)

0.748 1 **

(0.109)

0.685 2 ns (0.068)

0.811 1 ns (0.058)

0.739 2 ns (0.075)

0.807 NA (0.035)

CNS

0.717 Ref1 (0.139)

0.633 Ref2 (0.139)

0.7171 ns (0.075)

0.750 2 ns (0.059)

0.7331 ns (0.037)

0.750 2 ns (0.083)

0.867 1 ns (0.112)

0.767 2 ns (0.070)

0.722 NA (0.060)

Colon

0.826 Ref1 (0.097)

0.842 Ref2 (0.143)

0.8581 ns (0.081)

0.791 2 ns (0.071)

0.8261 ns (0.113)

0.792 2 ns (0.086)

0.844 1 ns (0.142)

0.760 2 * (0.089)

0.857 NA (0.055)

Leukemia

1.000 Ref1 (0.000)

0.943 Ref2 (0.060)

0.9581 ns (0.063)

0.930 2 ns (0.051)

0.9711 ns (0.064)

0.971 2 ns (0.064)

0.971 1 ns (0.064)

0.971 2 ns (0.064)

0.959 NA (0.025)

Lung

0.943 Ref1 (0.014)

0.935 Ref2 (0.021)

0.9431 ns (0.015)

0.951 2 ns (0.013)

0.9351 ns (0.010)

0.937 2 ns (0.026)

0.939 1 ns (0.016)

0.921 2 ns (0.006)

0.990 NA (0.009)

Lymphoma

1.000 Ref1 (0.000)

0.980 Ref2 (0.027)

0.9901 ns (0.023)

0.980 2 ns (0.027)

0.9901 ns (0.021)

0.971 2 ns (0.043)

1.000 1 NA (0.000)

0.970 2 ns (0.045)

0.977 NA (0.028)

MLL

0.953 Ref1 (0.048)

0.953 Ref2 (0.033)

0.9531 ns (0.058)

0.888 2 ns (0.104)

0.9711 ns (0.064)

0.962 2 ns (0.062)

0.953 1 ns (0.058)

0.925 2 ns (0.043)

0.943 NA (0.033)

SRBCT

0.994 Ref1 (0.013)

0.970 Ref2 (0.021)

0.9881 * (0.026)

0.982 2 * (0.026)

0.9691 ns (0.022)

0.988 2 ns (0.017)

0.988 1 ns (0.016)

0.939 2 ** (0.066)

0.992 NA (0.012)

  1. **p value of Wilcoxon signed-rank test is less than the significance level of 0.05
  2. *p value of Wilcoxon signed-rank test is less than the significance level of 0.1
  3. ns no significant