Sparse semi-supervised learning on low-rank kernel

doi:10.1016/j.neucom.2013.09.033

Sparse semi-supervised learning on low-rank kernel

2014-04-01

12

PRE

AI

DOI：10.1016/j.neucom.2013.09.033

Original

Original request for help

Save

Abstract

En 中文

Advances of modern science and engineering lead to unprecedented amount of data for information processing. Of particular interest is the semi-supervised learning, where very few training samples are available among large volumes of unlabeled data. Graph-based algorithms using Laplacian regularization have achieved state-of-the-art performance, but can induce huge memory and computational costs. In this paper, we introduce L-1-norm penalization on the low-rank factorized kernel for efficient, globally optimal model selection in graph-based semi-supervised learning. An important novelty is that our formulation can be transformed to a standard LASSO regression. On one hand, this makes it possible to employ advanced sparse solvers to handle large scale problems; on the other hand, a globally optimal subset of basis can be chosen adaptively given desired strength of penalizing model complexity, in contrast to some current endeavors that pre-determine the basis without coupling it with the learning task. Our algorithm performs competitively with state-of-the-art algorithms on a variety of benchmark data sets. In particular, it is orders of magnitude faster than exact algorithms and achieves a good trade-off between accuracy and scalability. (C) 2013 Elsevier B.V. All rights reserved.

Keywords:

Semi-supervised learning

Regularized least squares

Manifold regularization

Graph Laplacian

Sparse regression

Low-rank approximation

Sparse semi-supervised learning on low-rank kernel

Abstract

AI Summary

Journal

Organization