Deep Topic Models for Multilabel Learning
[edit]
Proceedings of Machine Learning Research, PMLR 89:28492857, 2019.
Abstract
We present a probabilistic framework for multilabel learning based on a deep generative model for the binary label vector associated with each observation. Our generative model learns deep multilayer latent embeddings of the binary label vector, which are conditioned on the input features of the observation. The model also has an interesting interpretation in terms of a deep topic model, with each label vector representing a bagofwords document, with the input features being its metadata. In addition to capturing the structural properties of the label space (e.g., a nearlowrank label matrix), the model also offers a clean, geometric interpretation. In particular, the nonlinear classification boundaries learned by the model can be seen as the union of multiple convex polytopes. Our model admits a simple and scalable inference via efficient Gibbs sampling or EM algorithm. We compare our model with stateoftheart baselines for multilabel learning on benchmark data sets, and also report some interesting qualitative results.
Related Material


