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 2006.09735stat.MLcs.DScs.LGmath.STstat.COstat.TH
收藏

基于截尾样本的稀疏图模型的有效统计

Efficient Statistics for Sparse Graphical Models from Truncated Samples

National University of Singapore(新加坡国立大学);Singapore University of Technology and Design(新加坡科技设计大学)

Arnab Bhattacharyya, Rathin Desai, Sai Ganesh Nagarajan, Ioannis Panageas

摘要:

本文研究截尾样本下的高维估计。我们专注于两个基本和经典的问题:(i)稀疏高斯图形模型的推理和(ii)支持稀疏线性模型的恢复。(i)对于高斯图形模型,假设$d$维样本${\bf x}$是从高斯$N(μ,Σ)$生成的,并且仅当它们属于子集$S \subseteq \mathbb{R}^d$时才被观察到。我们发现,$μ$和$Σ$可以估计错误$ε$在Frobenius范数,使用$\tilde{O}\left(\frac{\textrm{nz}(Σ^{-1})}{ε^2}\right)$样本从截断$\mathcal{N}(μ,Σ)$和访问$S$的成员预言。集合$S$被假定为在未知分布下具有非平凡测度,但在其他方面是任意的。(ii)对于稀疏线性回归,假设生成样本$({\bf x},y)$,其中仅当$y$属于截断集$S \subseteq \mathbb{R}$时才看到$y = {\bf x}^\top{Ω^*} + \mathcal{N}(0,1)$和$({\bf x}, y)$。我们考虑$Ω^*$是稀疏的情况下,支持集的大小$k$。我们的主要结果是建立精确的条件上的问题尺寸$d$,支持大小$k$,观测$n$的数量,和属性的样本和截断,足以恢复$Ω^*$的支持。具体而言,我们表明,在一些温和的假设下,只有$O(k^2 \log d)$样本需要估计$Ω^*$在$\ell_\infty$ -范数有界误差。对于这两个问题,我们的估计最小化的总和有限人口负对数似然函数和$\ell_1$正则化项。

英文摘要:

In this paper, we study high-dimensional estimation from truncated samples. We focus on two fundamental and classical problems: (i) inference of sparse Gaussian graphical models and (ii) support recovery of sparse linear models. (i) For Gaussian graphical models, suppose $d$-dimensional samples ${\bf x}$ are generated from a Gaussian $N(μ,Σ)$ and observed only if they belong to a subset $S \subseteq \mathbb{R}^d$. We show that $μ$ and $Σ$ can be estimated with error $ε$ in the Frobenius norm, using $\tilde{O}\left(\frac{\textrm{nz}(Σ^{-1})}{ε^2}\right)$ samples from a truncated $\mathcal{N}(μ,Σ)$ and having access to a membership oracle for $S$. The set $S$ is assumed to have non-trivial measure under the unknown distribution but is otherwise arbitrary. (ii) For sparse linear regression, suppose samples $({\bf x},y)$ are generated where $y = {\bf x}^\top{Ω^*} + \mathcal{N}(0,1)$ and $({\bf x}, y)$ is seen only if $y$ belongs to a truncation set $S \subseteq \mathbb{R}$. We consider the case that $Ω^*$ is sparse with a support set of size $k$. Our main result is to establish precise conditions on the problem dimension $d$, the support size $k$, the number of observations $n$, and properties of the samples and the truncation that are sufficient to recover the support of $Ω^*$. Specifically, we show that under some mild assumptions, only $O(k^2 \log d)$ samples are needed to estimate $Ω^*$ in the $\ell_\infty$-norm up to a bounded error. For both problems, our estimator minimizes the sum of the finite population negative log-likelihood function and an $\ell_1$-regularization term.

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。

↑