Research Article

Correlation-Based Ensemble Feature Selection Using Bioinspired Algorithms and Classification Using Backpropagation Neural Network

Table 2

Outline of WDBC dataset.

S. noFeatureDescriptionData type

1IDPatient identification numberNumerical
2DiagnosisMalignant or benignCharacter

3Radius (mean)Mean of distances from centre to points on the perimeterReal
4Radius (error)Real
5Radius (worst)Real

6Texture (mean)Standard deviation of grey scale valuesReal
7Texture (error)Real
8Texture (worst)Real

9Perimeter (mean)Perimeter of cell nucleusReal
10Perimeter (error)Real
11Perimeter (worst)Real

12Area (mean)Area of cellReal
13Area (error)Real
14Area (worst)Real

15Smoothness (mean)Local variation in radius lengthsReal
16Smoothness (error)Real
17Smoothness (worst)Real

18Compactness (mean)(perimeter^2/area—1.0)Real
19Compactness (error)Real
20Compactness (worst)Real

21Concavity (mean)Severity of concave portions of the contourReal
22Concavity (error)Real
23Concavity (worst)Real

24Concave (mean)Number of concave portions of the contourReal
25Concave (error)Real
26Concave (worst)Real

27Symmetry (mean)Measure of cell symmetryReal
28Symmetry (error)Real
29Symmetry (worst)Real

30Fractal dimension (mean)(“Coastline approximation”—1)Real
31Fractal dimension (error)Real
32Fractal dimension (worst)Real