Recovery Guarantee of Non-negative Matrix Factorization via Alternating Updates
Li, Liang, Risteski · cs.LG,cs.DS,stat.ML · 2016-11-11 · 原文
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.
讲义
讲义·推断 依据「原文」自动生成的结构化摘要（推断），非原文表述；以原文为准。
1. 人话版
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.
2. 领域脉络
本文类目：cs.LG、cs.DS、stat.ML，属于其所在研究脉络的最新进展。
3. 机制拆解
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.
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.
4. 证据与数字
摘要未给出量化结果——留意原文的实验与数据。
5. 反例与边界
In particular, its only essential requirement on features is linear independence.
6. 跨领域连接与意外收获
横跨 3 个类目（cs.LG、cs.DS、stat.ML），关注其在你兴趣板块间的迁移面。
7. 可复用方法
把本文机制与你手头项目对照，找一个两周内能验证的最小实验。
8. 术语表
精读时把不熟的术语记入此处，作为下次回忆的锚点。