Recovery Guarantee of Non-negative Matrix Factorization via Alternating Updates

Abstract

Non-negative matrix factorization is a popular tool for decomposing data into feature and weight matrices under non-negativity constraints. It enjoys practical success but is poorly understood theoretically. This paper proposes an algorithm that alternates between decoding the weights and updating the features, and shows that assuming a generative model of the data, it provably recovers the ground-truth under fairly mild conditions. In particular, its only essential requirement on features is linear independence. Furthermore, the algorithm uses ReLU to exploit the non-negativity for decoding the weights, and thus can tolerate adversarial noise that can potentially be as large as the signal, and can tolerate unbiased noise much larger than the signal. The analysis relies on a carefully designed coupling between two potential functions, which we believe is of independent interest.

arXiv commentTo appear in NIPS 2016. 8 pages of extended abstract; 48 pages in total

References in corpus (3)

• Simple, Efficient, and Neural Algorithms for Sparse Coding
• A Hebbian/Anti-Hebbian Network Derived from Online Non-Negative Matrix Factorization Can Cluster and Discover Sparse Features
• A Normative Theory of Adaptive Dimensionality Reduction in Neural Networks

Cited by in corpus (9)

• Towards Understanding Ensemble, Knowledge Distillation and Self-Distillation in Deep Learning
• Manipulation of Granular Materials by Learning Particle Interactions
• Consistent Estimation of Mixed Memberships with Successive Projections
• Provable quantum state tomography via non-convex methods
• Recovery and Generalization in Over-Realized Dictionary Learning
• Forward Super-Resolution: How Can GANs Learn Hierarchical Generative Models for Real-World Distributions
• Understanding Notions of Stationarity in Non-Smooth Optimization
• Making Method of Moments Great Again? -- How can GANs learn distributions
• Crowdsourcing via Annotator Co-occurrence Imputation and Provable Symmetric Nonnegative Matrix Factorization