The game theoretic p-Laplacian and semi-supervised learning with few labels

Re-assign community