Abstract

One of the most challenging machine learning problems is a particular case of data classification in which classes are hierarchically structure and objects can be assigned to multiple paths of the class hierarchy at the same time. This task is known as hierarchical multi-label classification (HMC), with applications in text classification, image annotation, and in bioinformatics problems such as protein function prediction. In this paper we propose novel neural network architectures for HMC called HMCN, capable of simultaneously optimizing local and global loss functions for dis covering local hierarchical class-relationships an global information from the entire class hierarchy while penalizing hierarchical violations. We evaluate its performance in 21 datasets from four distinct domains, and we compare it against the current HMC state-of-the-art approaches. Results show that HMCN substantially outperforms all baselines with statistical significance, arising a the novel state-of-the-art for HMC.