Dual Relation Semi-Supervised Multi-Label Learning

Multi-label learning problems are commonly found in many applications. A characteristic shared by many multi-label learning problems is that some labels have significant correlations between them. In this paper, we propose a novel multi-label learning method, called multi-label relationship learning (MLRL), which extends the conventional support vector machine by explicitly learning and utilizing the relationships between labels. Specifically, we model the label relationships using a label covariance matrix and use it to define a new regularization term for the optimization problem. MLRL learns the model parameters and the label covariance matrix simultaneously based on a unified convex formulation. To solve the convex optimization problem, we use an alternating method in which each subproblem can be solved efficiently. The relationship between MLRL and two widely used maximum margin methods for multi-label learning is investigated. Moreover, we also propose a semi-supervised extension of MLRL, called SSMLRL, to demonstrate how to make use of unlabeled data to help learn the label covariance matrix. Through experiments conducted on some multi-label applications, we find that MLRL not only gives higher classification accuracy but also has better interpretability as revealed by the label covariance matrix.