Rank-one matrix estimation: analysis of algorithmic and information theoretic limits by the spatial coupling method

Abstract

Factorizing low-rank matrices is a problem with many applications in machine learning and statistics, ranging from sparse PCA to community detection and sub-matrix localization. For probabilistic models in the Bayes optimal setting, general expressions for the mutual information have been proposed using powerful heuristic statistical physics computations via the replica and cavity methods, and proven in few specific cases by a variety of methods. Here, we use the spatial coupling methodology developed in the framework of error correcting codes, to rigorously derive the mutual information for the symmetric rank-one case. We characterize the detectability phase transitions in a large set of estimation problems, where we show that there exists a gap between what currently known polynomial algorithms (in particular spectral methods and approximate message-passing) can do and what is expected information theoretically. Moreover, we show that the computational gap vanishes for the proposed spatially coupled model, a promising feature with many possible applications. Our proof technique has an interest on its own and exploits three essential ingredients: the interpolation method first introduced in statistical physics, the analysis of approximate message-passing algorithms first introduced in compressive sensing, and the theory of threshold saturation for spatially coupled systems first developed in coding theory. Our approach is very generic and can be applied to many other open problems in statistical estimation where heuristic statistical physics predictions are available.

arXiv commentSubmitted to Journal of Machine Learning Research (JMLR)

References in corpus (12)

• Stochastic blockmodels and community structure in networks
• Sparse Principal Components Analysis
• Statistical-Computational Tradeoffs in Planted Problems and Submatrix Localization with a Growing Number of Clusters and Submatrices
• Matrix Completion from a Few Entries
• MMSE of probabilistic low-rank matrix estimation: Universality with respect to the output channel
• Mutual Information in Rank-One Matrix Estimation
• Asymptotic Mutual Information for the Two-Groups Stochastic Block Model
• Finite-range spin glasses in the Kac limit: free energy and local observables
• Spectral Clustering of Graphs with the Bethe Hessian
• Submatrix localization via message passing
• Dynamics and termination cost of spatially coupled mean-field models
• Matrix Completion from Fewer Entries: Spectral Detectability and Rank Estimation

Cited by in corpus (4)

• The generalization error of max-margin linear classifiers: Benign overfitting and high dimensional asymptotics in the overparametrized regime
• Statistical limits of dictionary learning: random matrix theory and the spectral replica method
• Mutual Information for the Stochastic Block Model by the Adaptive Interpolation Method
• Gauge theory for mixed $p$-spin glasses