Related work | For prediction | For description | Generalization of tree structure | Number of trees | Primary-feature discovery | Feature selection |
---|---|---|---|---|---|---|
SON ET AL. [7] | yes | yes | tree from feature selection | 1 | rough set attribute reduced on 10-fold cross-validation | |
STIGLIC ET AL. [12] | yes | yes | tuning the tree fitting in one screen | 1 | ||
SCHEURWEGS ET AL. [15] | yes | many | selecting primary features using the internal scoring metric in Random Forest | |||
BREIMAN [16] | yes | many | ||||
JOLOUDARI ET AL. [18] | yes | yes | rules selecting from parts of trees | many | ranking of predictor significant | |
MOHAN ET AL. [20] | yes | many | apriori algorithm | |||
GHOSH ET AL. [21] | yes | many | Relief and LASSO | |||
ASHRI ET AL. [22] | yes | many | genetic algorithm | |||
MORENO-SANCHEZ [23] | yes | yes | decision tree constructed from feature selection at the maximum level 3 | many for prediction, 1 for description | feature-important measure | |
A DESCRIPTIVE FOREST | yes | the least PTS | many | association-rule tree with a constraining rule |