Ordinal Trees and Random Forests: Score-Free Recursive Partitioning and Improved Ensembles

Abstract Existing ordinal trees and random forests typically use scores that are assigned to the ordered categories, which implies a higher scale level is used. Versions of proposed take seriously avoid assignment artificial scores. The construction principle based on an investigation binary models implicitly used in parametric regression. These building blocks can be fitted by combined similar way as models. obtained only. Since constituent elements trees, one exploit wide range have already been developed. A further topic potentially poor performance forests, seems neglected literature. Ensembles include obtain prediction methods tend perform well settings. evaluated empirically using several data sets.