Sparse semi-supervised learning on low-rank kernel

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 a...