XGBoost parameters

The parameters accepted by the algorithm include:

  • objective to define the type of task, say regression or classification;
  • colsample_bytree the subsample ratio of columns when constructing each tree. Subsampling happens once in every iteration. This number is usually a value between 0 and 1;
  • learning_rate that determines how fast or slow the model will learn;
  • max_depth indicates the maximum depth for each tree. The more the trees, the greater model complexity, and the higher chances of overfitting;
  • alpha is the L1 regularization on weights;
  • n_estimators is the number of decision trees to fit.