Table 1
Default and Bayesian optimized parameters for the random forest and gradient boosting algorithms, learning rate is not a hyperparameter of the random forest algorithm, minimum impurity decrease was also optimized but did not show a change from 0.0, variable max depth of the default random forest is defined as the expansion of tree until endpoints contain less than [min_samples_split] samples.
Algorithm default/optimized | # Trees | Max features per tree | Min samples split | Min samples leaf | Max depth | Learning rate |
---|---|---|---|---|---|---|
Random Forest | 100/800 | 6/3 | 2/2 | 1/1 | var./11 | N.A. |
Gradient Boost | 100/490 | 6/3 | 2/20 | 1/8 | 3/5 | 0.1/0.02 |
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.