Dynamic construction of Random Forests: Evaluation using biomedical engineering problems

The aim of this work is the development of a method for the automatic determination of the optimum number of base classifiers which consists of the Random Forests. The novelty of the proposed method is that it doesn't need to select the classifiers to be in the final ensemble from a pool of classifiers which is known in advance, but determines the number of classifiers dynamically during the growing procedure of the forest. The method is based on the employment of an online fitting procedure and is evaluated using classical Random Forests and its modifications as ensemble methods.