Abstract:Conditional random field (CRFs) is a popular and effective approach to structured prediction. When the underlying structure does not have a small tree-width, maximum likelihood estimation (MLE) is in general com- putationally hard. Discriminative methods such as Perceptron or Max-Margin Markov Networks circumvent this problem by requiring the MAP assignment only, which is often more tractable, either exactly or approx- imately with linear programming (LP) relaxations. In this paper, we propose an approximate learning method for MLE of CRFs. We leverage LP relaxations to find multiple diverse MAP solutions and use them to approx- imate the intractable partition function. The proposed approach is easy to parallelize, and yields competitive performance in test accuracies on several structured prediction tasks.
Enter your feedback below and we'll get back to you as soon as possible.