From: Social networks for enhanced player churn prediction in mobile free-to-play games
Classifier | Hyper parameters | Candidate setting |
---|---|---|
DecisionTree | Alpha | 0.1, 0.01, 0.001, 0.0001 |
Max tree depth | 4, 8, 10, 20, 30 | |
Criterion | gini, entropy | |
KNN | Weights | uniform, distance |
# Neighbors | 3,5,11,14, 16, 18, 20 | |
Metric | euclidean, manhattan | |
Algorithm | auto,ball_tree, kd_tree, brute | |
RandomForest | # Estimators | 300, 500, 700 |
Max tree depth | 4, 8, 10, 20 | |
Criterion | gini, entropy | |
XGBoost | Minimum child weight | 1, 5, 10 |
Gamma | 1, 0.01, 0.1, 0.5 | |
Subsample ratio of the training instances | 0.6, 1.0 | |
Subsample ratio of columns | 0.6, 1.0 | |
Max tree depth | 3, 3, 5 | |
SGD | Loss | hinge, log, squared_hinge, modified_huber |
Alpha | 0.001, 0.01, 0.1 | |
Penalty | l2, l1, none |