Multi-label classification using conditional dependency networks
Yuhong Guo, Suicheng Gu · 2011
In this paper, we tackle the challenges of multi-label classification by developing a general condi-tional dependency network model. The proposed model is a cyclic directed graphical model, which provides an intuitive representation for the depen-dencies among multiple label variables, and a well integrated framework for efficient model training using binary classifiers and label predictions using Gibbs sampling inference. Our experiments show the proposed conditional model can effectively ex-ploit the label dependency to improve multi-label classification performance. 1