Skip to main content

Table 8 Random Forests feature selection aggregate ranking

From: Machine learning can predict survival of patients with heart failure from serum creatinine and ejection fraction alone

Final rankFeatureAccuracy decreaseAccuracy decrease rankGini impurityGini impurity rank
1Serum creatinine3.78×10−2111.841
2Ejection fraction3.43×10−2210.712
3Age1.53×10−238.583
4Creatinine phosphokinase7.27×10−467.264
4Serum sodium7.20×10−346.496
6Sex1.64×10−351.128
6Platelets2.47×10−486.805
8High blood pressure−1.68×10−3111.137
8Smoking3.68×10−470.9511
10Anaemia−5.91×10−4101.069
10Diabetes−1.41×10−491.0210
  1. We merged the two rankings through their position, through the Borda’s method [103]