Noise-robust semi-supervised learning via fast sparse coding

• We propose novel semi-supervised learning based on fast sparse coding.
• Our algorithm achieves promising results in noise-robust image classification.
• Our algorithm can readily be extended to many other challenging problems.

This paper presents a novel noise-robust graph-based semi-supervised learning algorithm to deal with the challenging problem of semi-supervised learning with noisy initial labels. Inspired by the successful use of sparse coding for noise reduction, we choose to give new L1-norm formulation of Laplacian regularization for graph-based semi-supervised learning. Since our L1-norm Laplacian regularization is explicitly defined over the eigenvectors of the normalized Laplacian matrix, we formulate graph-based semi-supervised learning as an L1-norm linear reconstruction problem which can be efficiently solved by sparse coding. Furthermore, by working with only a small subset of eigenvectors, we develop a fast sparse coding algorithm for our L1-norm semi-supervised learning. Finally, we evaluate the proposed algorithm in noise-robust image classification. The experimental results on several benchmark datasets demonstrate the promising performance of the proposed algorithm.