Skip to yearly menu bar Skip to main content


Convex Two-Layer Modeling with Latent Structure

Vignesh Ganapathiraman · Xinhua Zhang · Yaoliang Yu · Junfeng Wen

Area 5+6+7+8 #47

Keywords: [ Structured Prediction ] [ (Other) Classification ]


Unsupervised learning of structured predictors has been a long standing pursuit in machine learning. Recently a conditional random field auto-encoder has been proposed in a two-layer setting, allowing latent structured representation to be automatically inferred. Aside from being nonconvex, it also requires the demanding inference of normalization. In this paper, we develop a convex relaxation of two-layer conditional model which captures latent structure and estimates model parameters, jointly and optimally. We further expand its applicability by resorting to a weaker form of inference---maximum a-posteriori. The flexibility of the model is demonstrated on two structures based on total unimodularity---graph matching and linear chain. Experimental results confirm the promise of the method.

Live content is unavailable. Log in and register to view live content