Hierarchical Multilabel Classification Using Top-Down Label Combination and Artificial Neural Networks

Hierarchical Multilabel Classification is a classification problem where the classes of the examples are hierarchically structured and, additionally, each example can simultaneously belong to two or more classes in the same hierarchical level. This paper proposes a new Top-Down classification method based on a label combination process, using Artificial Neural Networks as base classifiers. The experimental evaluation used Bioinformatics datasets, and showed that the proposed method achieved good results in comparison with well-known methods from the literature.