Matrix Completion with Model-free Weighting

TL;DR

This paper introduces a model-free weighting approach for matrix completion that effectively handles non-uniform missing data, providing strong theoretical guarantees and demonstrating superior performance in heterogeneous missing scenarios.

Contribution

The paper presents a novel, computationally efficient weighting method for matrix completion that does not require explicit modeling of observation probabilities, with improved theoretical guarantees.

Findings

Achieves stronger guarantees under heterogeneous missing data.

Demonstrates effectiveness through numerical experiments.

Provides a new minimax lower bound for heterogeneous settings.

Abstract

In this paper, we propose a novel method for matrix completion under general non-uniform missing structures. By controlling an upper bound of a novel balancing error, we construct weights that can actively adjust for the non-uniformity in the empirical risk without explicitly modeling the observation probabilities, and can be computed efficiently via convex optimization. The recovered matrix based on the proposed weighted empirical risk enjoys appealing theoretical guarantees. In particular, the proposed method achieves a stronger guarantee than existing work in terms of the scaling with respect to the observation probabilities, under asymptotically heterogeneous missing settings (where entry-wise observation probabilities can be of different orders). These settings can be regarded as a better theoretical model of missing patterns with highly varying probabilities. We also provide a new minimax lower bound under a class of heterogeneous settings. Numerical experiments are also provided to demonstrate the effectiveness of the proposed method.