Multi-Scale Label Relation Learning for Multi-Label Classification Using 1-Dimensional Convolutional Neural Networks
This addresses the issue of under-/over-fitting in multi-label classification for researchers and practitioners by offering a more efficient alternative to RNN-based models, though it is incremental as it builds on existing convolutional approaches.
The paper tackles the problem of high parameter counts in recurrent neural network (RNN) models for multi-label classification by proposing Multi-Scale Label Dependence Relation Networks (MSDN), which uses 1-dimensional convolutional neural networks (1D-CNNs) to learn label dependencies at multiple scales, achieving better accuracies with drastically fewer parameters on public benchmark datasets.
We present Multi-Scale Label Dependence Relation Networks (MSDN), a novel approach to multi-label classification (MLC) using 1-dimensional convolution kernels to learn label dependencies at multi-scale. Modern multi-label classifiers have been adopting recurrent neural networks (RNNs) as a memory structure to capture and exploit label dependency relations. The RNN-based MLC models however tend to introduce a very large number of parameters that may cause under-/over-fitting problems. The proposed method uses the 1-dimensional convolutional neural network (1D-CNN) to serve the same purpose in a more efficient manner. By training a model with multiple kernel sizes, the method is able to learn the dependency relations among labels at multiple scales, while it uses a drastically smaller number of parameters. With public benchmark datasets, we demonstrate that our model can achieve better accuracies with much smaller number of model parameters compared to RNN-based MLC models.