arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2111.00327cs.ITmath.ITmath.STstat.MLstat.TH
收藏

超越独立测量:具有GNN应用的一般压缩感知

Beyond Independent Measurements: General Compressed Sensing with GNN Application

Alireza Naderi, Yaniv Plan

摘要:

我们考虑的问题,恢复一个结构化的信号$\mathbf{x} \in \mathbb{R}^{n}$从嘈杂的线性观测$\mathbf{y} =\mathbf{M} \mathbf{x}+\mathbf{w}$。测量矩阵被建模为$\mathbf{M} = \mathbf{B}\mathbf{A}$,其中$\mathbf{B} \in \mathbb{R}^{l \times m}$是任意的,并且$\mathbf{A} \in \mathbb{R}^{m \times n}$具有独立的子高斯行。通过改变$\mathbf{B}$和$\mathbf{A}$的亚高斯分布,这给出了一系列测量矩阵,这些测量矩阵可能具有重尾、相关的行和列以及具有大动态范围的奇异值。当结构是作为一个可能的非凸锥$T \subset \mathbb{R}^{n}$,一个近似的经验风险最小化被证明是一个强大的估计,如果有效的测量数是足够的,即使在存在模型失配。在具有独立(子)高斯测量的经典压缩感知中,人们会问需要多少测量来恢复$\mathbf{x}$?然而,在我们的设置中,有效的测量次数取决于$\mathbf{B}$的属性。我们表明,$\mathbf{B}$的有效秩可以用作测量次数的替代品,如果这超过了$(T-T) \cap \mathbb{S}^{n-1}$的高斯平均宽度的平方,则可以保证准确的恢复。此外,我们详细研究了生成先验的特殊情况,即当$\mathbf{x}$接近$T = \mathrm{ran}(G)$并且$G: \mathbb{R}^k \rightarrow \mathbb{R}^n$是具有ReLU激活函数的生成神经网络(GNN)时。我们的工作依赖于Jeong,Li,Plan和Yilmaz arXiv:2001.10631在随机矩阵理论中的最新结果。.

英文摘要:

We consider the problem of recovering a structured signal $\mathbf{x} \in \mathbb{R}^{n}$ from noisy linear observations $\mathbf{y} =\mathbf{M} \mathbf{x}+\mathbf{w}$. The measurement matrix is modeled as $\mathbf{M} = \mathbf{B}\mathbf{A}$, where $\mathbf{B} \in \mathbb{R}^{l \times m}$ is arbitrary and $\mathbf{A} \in \mathbb{R}^{m \times n}$ has independent sub-gaussian rows. By varying $\mathbf{B}$, and the sub-gaussian distribution of $\mathbf{A}$, this gives a family of measurement matrices which may have heavy tails, dependent rows and columns, and singular values with a large dynamic range. When the structure is given as a possibly non-convex cone $T \subset \mathbb{R}^{n}$, an approximate empirical risk minimizer is proven to be a robust estimator if the effective number of measurements is sufficient, even in the presence of a model mismatch. In classical compressed sensing with independent (sub-)gaussian measurements, one asks how many measurements are needed to recover $\mathbf{x}$? In our setting, however, the effective number of measurements depends on the properties of $\mathbf{B}$. We show that the effective rank of $\mathbf{B}$ may be used as a surrogate for the number of measurements, and if this exceeds the squared Gaussian mean width of $(T-T) \cap \mathbb{S}^{n-1}$, then accurate recovery is guaranteed. Furthermore, we examine the special case of generative priors in detail, that is when $\mathbf{x}$ lies close to $T = \mathrm{ran}(G)$ and $G: \mathbb{R}^k \rightarrow \mathbb{R}^n$ is a Generative Neural Network (GNN) with ReLU activation functions. Our work relies on a recent result in random matrix theory by Jeong, Li, Plan, and Yilmaz arXiv:2001.10631. .

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑