Skip to main content

Table 3 Overview of precision and recall of all analyzed feature selection methods for different simulation settings

From: Cost-Constrained feature selection in binary classification: adaptations for greedy forward selection and genetic algorithms

 Recall Precision
 ABCDEFGHKABCDEFGHK
Methods
FS33.958.319.253.779.434.49.724.018.998.798.398.8100.069.993.959.191.298.1
cFS41.666.729.162.769.340.220.527.928.295.5100.092.593.137.968.258.785.792.3
cFS.mean38.865.725.259.378.340.318.627.024.797.899.097.497.859.889.965.892.097.5
cFS.max37.664.023.657.079.138.716.326.322.998.799.098.698.863.591.165.693.097.9
fGA40.666.327.455.081.738.019.027.823.897.999.598.187.359.383.768.193.898.3
cGA40.366.327.361.779.938.619.027.823.897.899.597.096.961.290.069.794.297.4
Filter.tTest34.358.718.453.098.434.78.823.918.999.599.499.8100.088.295.661.593.699.8
Filter.Symuncert24.156.018.447.084.233.08.724.118.699.3100.099.697.284.095.057.999.399.8
Filter.PraznikJMIM30.357.318.644.060.828.17.023.818.483.186.097.676.354.477.043.389.396.0
Filter.RangerImpurity34.357.019.153.796.033.48.924.118.893.585.599.098.885.093.360.589.998.3
Reference
Budget constraint50.066.733.366.7100.050.033.333.333.3100.0100.0100.0100.0100.0100.0100.0100.0100.0
Random selection20.54.71.50.31.10.51.64.01.660.010.010.01.01.01.310.010.010.0
  1. Values are given in percent. The results are an extended version of the data shown in Fig. 4. Please refer to the description of this figure for further details