RIPML: A Restricted Isometry Property based Approach to Multilabel Learning
This addresses multilabel learning for large-scale datasets with sparse labels, but it is incremental as it builds on existing linear dimensionality reduction methods.
The paper tackles multilabel learning with extreme label sparsity by proposing RIPML, a method that projects labels onto a random low-dimensional subspace using the Restricted Isometry Property and uses k-nearest neighbor for inference, achieving results comparable to state-of-the-art linear dimensionality reduction approaches in simulations.
The multilabel learning problem with large number of labels, features, and data-points has generated a tremendous interest recently. A recurring theme of these problems is that only a few labels are active in any given datapoint as compared to the total number of labels. However, only a small number of existing work take direct advantage of this inherent extreme sparsity in the label space. By the virtue of Restricted Isometry Property (RIP), satisfied by many random ensembles, we propose a novel procedure for multilabel learning known as RIPML. During the training phase, in RIPML, labels are projected onto a random low-dimensional subspace followed by solving a least-square problem in this subspace. Inference is done by a k-nearest neighbor (kNN) based approach. We demonstrate the effectiveness of RIPML by conducting extensive simulations and comparing results with the state-of-the-art linear dimensionality reduction based approaches.