Friend Recommender System for Social Networks Based on Stacking Technique and Evolutionary Algorithm
Table 3
Some of important parameters considered for different methods.
Method
Parameters
XGBoost
Number of boosting stages = 100 loss function = “log_loss” Learning rate = 0.1 Maximum depth of the individual regression estimators = 3
RandomForest_1
The number of trees in the forest = 20 measures the quality of a split = “gini” The number of features to consider when looking for the best split = “sqrt (n_features)”
RandomForest_2
The number of trees in the forest = 30 measures the quality of a split = “gini” The number of features to consider when looking for the best split = “sqrt (num of features)”
Logistic regression
Solver = “lbfgs” penalty term = “L2” Tolerance for stopping criteria = 1e-4