RF |
criterion |
measurement methods of impurities |
gini |
|
n_estimators |
number of trees in the forest |
110 |
|
max_depth |
the maximum depth of the tree |
5 |
|
max_features |
maximum number of retained features when branching |
10 |
GBDT |
n_estimators |
number of trees in the forest |
45 |
|
max_depth |
the maximum depth of the tree |
2 |
|
min_samples_split |
the number of training samples at least included
in the sub-node |
2 |
|
learning_rate |
Learning rate |
0.1 |
Decision Tree |
criterion |
measurement methods of impurities |
entropy |
|
random_state |
parameters of random patterns in branches |
13 |
|
splitter |
way of the tree branches |
random |
|
max_depth |
the maximum depth of the tree |
5 |
KNN |
n_neighbors |
number of nearest neighbor samples with voting
rights |
5 |
|
weights |
voting proportion method |
uniform |
Naïve Bayes |
alpha |
Laplacian smoothing coefficient |
20 |