Manifold Regularization
Lorenzo Rosasco
Description
We first analyze the limits of learning in high dimension. Hence, we stress
the difference between high dimensional ambient space and intrinsic geometry associated to the
marginal distribution. We observe that, in the
semi-supervised setting, unlabeled data could be used to exploit low dimensionality of
the intrinsic geometry. In order to formalize these
intuitions we briefly introduce the manifold Laplacian and Graph Laplacian. Finally,
we introduce a new class of regularization algorithms, aimed at
enforcing smoothness relative to the intrinsic geometry.
Slides
Slides for this lecture: PDF
Suggested Reading