Consistent Semi-Supervised Graph Regularization for High Dimensional Data

Semi-supervised Laplacian regularization, a standard graph-based approach for learning from both labelled and unlabelled data, was recently demonstrated to have an insignificant high dimensional learning efficiency with respect to unlabelled data (Mai and Couillet, 2018), causing it to be outperformed by its unsupervised counterpart, spectral clustering, given sufficient unlabelled data. Following a detailed discussion on the origin of this inconsistency problem, a novel regularization approach involving centering operation is proposed as solution, supported by both theoretical analysis and empirical results.

Automated summary

A new regularization approach involving centering operation is proposed as a solution to the high-dimensional learning efficiency problem in semi-supervised learning, supported by both theoretical analysis and empirical results.