Table A2.

Classifier parameter definition.

Classifier	Parameter	Parameter Explanation
KNN	weights = ‘distance’, p = 1, n_neighbors = 6, leaf_size = 2, algorithm = ‘ball_tree’	Weights: weight function used in prediction; p: power parameter for the Minkowski metric; n_neighbors: number of neighbors to use; leaf_size: leaf size passed to BallTree; algorithm: used to compute the nearest neighbors
RF	n_estimators = 90, oob_score = True, random_state = 10	n_estimators: the number of trees in the forest; oob_score: whether to use out-of-bag samples; random_state: controls both the randomness of the bootstrapping of the samples used when building trees and the sampling of the features to consider when looking for the best split at each node
DT	criterion = ‘gini’, max_depth = 6, splitter = ‘best’	criterion: the function to measure the quality of a split; max_depth: the maximum depth of the tree; splitter: the strategy used to choose the split at each node
GBDT	n_estimators = 120, max_depth = 10, learning_rate = 0.01, min_samples_split = 4, subsample = 0.5	n_estimators: the number of boosting stages to perform; max_depth: maximum depth of the regression estimators; learning_rate:learning rate shrinks the contribution of each tree by learning_rate; min_samples_split: the minimum number of samples required to split an internal node; subsample: the fraction of samples to be used for fitting the individual base learners
AdaBoost	n_estimators = 6, learning_rate = 0.1	n_estimators: the maximum number of estimators at which boosting is terminated; learning_rate: learning rate shrinks the contribution of each classifier