Algorithm 1.

Weighted Random Forest algorithm pseudocode.

input: Number of Trees (Ntree), random subset of the features (mtry), training dataset (D)

output: Random Forest (RF)

1: RF is empty

2: for each j to Ntree do

3: Di = Bootstrap Sample (D)

4: DTi = Random Decision Tree (Di, mtry)

5: RF = RF ∪ DTi

6: end

7: for each i to n do

8: Compute $w_i^{obs}$ using Formula (5)

9: end

10: for each j to Ntree do

11: ${\theta }_j=f\left(AUC{\left(w^{obs}\right)}_{IN{B}_j},AUC{\left(w^{obs}\right)}_{OO{B}_j}\right)$

12: end

13: for each j to Ntree do

14: Compute $w_j$ using Formula (9)

15: end

16: for each i to n do

17: Compute final prediction ${\widehat{\mathit{\Upsilon }}}_i^{RF}$ using Formula (3)

18: end

19: return RF