Improving Local Per Level Hierarchical Classification

In the domain of many relevant classification problems, classes are organized in hierarchies, representing specialization relationships between them. These are the so-called hierarchical classification problems. Methods based on different approaches have been used to solve them, trying to achieve better predictive performance. In this work, we propose two local per level hierarchical classifiers, which contain distinct strategies to solve inconsistent predictions, common to the local per level approach. We have compared the proposed methods with traditional strategies from different paradigms. The computational experiments, conducted over 18 hierarchical classification data sets, showed that the proposed ideas were able to reach competitive and robust results in terms of prediction accuracy.