Defining the Ranges Boundaries of the Optimal Parameters Values for the Random Forest Classifier

The problem of determining the search ranges for the optimal parameters values of the random forest classifier with the aim of reducing the time spent on its development has been considered. The recommendations for determining the boundaries ranges for the such parameters of the random forest classifier as the number of trees, the number of features for splitting, the depth of trees and the minimum number of objects at the leaf have been formed on the basis of the experimental results obtained during the development of the random forest classifiers using the different datasets from the machine learning data repositories. The results of experimental studies on the development of the random forest classifiers using the training and test sets based on the analyzed datasets are given, the graphical dependences on the classification quality estimation at the test sets and the development time of the random forest classifiers at the training sets are obtained in general form.