It is well-known that exploiting label correlations is important tomulti-label learning. Existing approaches either assume that the labelcorrelations are global and shared by all instances; or that the labelcorrelations are local and shared only by a data subset. In fact, in thereal-world applications, both cases may occur that some label correlations areglobally applicable and some are shared only in a local group of instances.Moreover, it is also a usual case that only partial labels are observed, whichmakes the exploitation of the label correlations much more difficult. That is,it is hard to estimate the label correlations when many labels are absent. Inthis paper, we propose a new multi-label approach GLOCAL dealing with both thefull-label and the missing-label cases, exploiting global and local labelcorrelations simultaneously, through learning a latent label representation andoptimizing label manifolds. The extensive experimental studies validate theeffectiveness of our approach on both full-label and missing-label data.
展开▼
