Generalization error bounds for kernel matrix completion and extrapolation

Abstract

Prior information can be incorporated in matrix completion to improve estimation accuracy and extrapolate the missing entries. Reproducing kernel Hilbert spaces provide tools to leverage the said prior information, and derive more reliable algorithms. This paper analyzes the generalization error of such approaches, and presents numerical tests confirming the theoretical results.

References in corpus (8)

• Signal Recovery on Graphs: Variation Minimization
• Matrix Completion on Graphs
• Provable Inductive Matrix Completion
• Low-rank matrix factorization with attributes
• Rank regularization and Bayesian inference for tensor completion and extrapolation
• Concentration-Based Guarantees for Low-Rank Matrix Reconstruction
• Transductive Rademacher Complexity and its Applications
• Matrix completion and extrapolation via kernel regression

Cited by in corpus (1)

• Orthogonal Inductive Matrix Completion