Consistency of Lipschitz learning with infinite unlabeled data and finite labeled data
This addresses a theoretical problem in semi-supervised learning for researchers, showing incremental insights into the behavior of Lipschitz learning under different graph weight models.
The paper tackles the consistency of Lipschitz learning on graphs with infinite unlabeled and finite labeled data, proving that it is insensitive to unlabeled data distribution in a random geometric graph with kernel-based weights but highly sensitive with self-tuning weights, where sensitivity can be adjusted by tuning weights.
We study the consistency of Lipschitz learning on graphs in the limit of infinite unlabeled data and finite labeled data. Previous work has conjectured that Lipschitz learning is well-posed in this limit, but is insensitive to the distribution of the unlabeled data, which is undesirable for semi-supervised learning. We first prove that this conjecture is true in the special case of a random geometric graph model with kernel-based weights. Then we go on to show that on a random geometric graph with self-tuning weights, Lipschitz learning is in fact highly sensitive to the distribution of the unlabeled data, and we show how the degree of sensitivity can be adjusted by tuning the weights. In both cases, our results follow from showing that the sequence of learned functions converges to the viscosity solution of an $\infty$-Laplace type equation, and studying the structure of the limiting equation.