1 paper · 1 filter
Mauricio Flores, Jeff Calder, Gilad Lerman
This paper addresses theory and applications of ℓp-based Laplacian regularization in semi-supervised learning. The graph p-Laplacian for p>2 has been proposed recently as…