Open Access

Table B.1.

Grids used in the hyper-parameter optimization procedure.

Algorithm Default hyper-parameters Optimization grid
LightGBM n_estimators = 100 learning_rate = 0.1 max_depth = -1 n_estimators = {50,100,250,500,750,1000} learning_rate = {0.5,0.3,0.1,0.075,0.05,0.02,0.01} max_depth = {-1,4,5,6,7,8}

XGBoost n_estimators = 100 learning_rate = 0.3 max_depth = 6 n_estimators = {50,100,250,500,1000,1500,2000} learning_rate = {0.5,0.3,0.2,0.1,0.075,0.05,0.02,0.01} max_depth = {4,5,6,7,8}

CatBoost n_estimators = 1000 learning_rate = 0.03 max_depth = 6 n_estimators = {250,500,600,700,800,900,1000,1250,2000} learning_rate = {0.5,0.3,0.1,0.075,0.05,0.03,0.01} max_depth = {4,5,6,7,8}

Notes. The values for max_depth of 0 or -1 result in not limited trees.

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.