We study the Kronecker product regression problem, in which the design matrix is a Kronecker product of two or more matrices. Given $A_i \in \mathbb{R}^{n_i \times d_i}$ for $i=1,2,\dots,q$ where $n_i \gg d_i$ for each $i$, and $b \in \mathbb{R}^{n_1 n_2 \cdots n_q}$, let $\mathcal{A} = A_1 \otimes A_2 \otimes \cdots \otimes A_q$. Then for $p \in [1,2]$, the goal is to find $x \in \mathbb{R}^{d_1 \cdots d_q}$ that approximately minimizes $\|\mathcal{A}x - b\|_p$. Recently, Diao, Song, Sun, and Woodruff (AISTATS, 2018) gave an algorithm which is faster than forming the Kronecker product $\mathcal{A}$ Specifically, for $p=2$ their running time is $O(\sum_{i=1}^q \text{nnz}(A_i) + \text{nnz}(b))$, where nnz$(A_i)$ is the number of non-zero entries in $A_i$. Note that nnz$(b)$ can be as large as $n_1 \cdots n_q$. For $p=1,$ $q=2$ and $n_1 = n_2$, they achieve a worse bound of $O(n_1^{3/2} \text{poly}(d_1d_2) + \text{nnz}(b))$. In this work, we provide significantly faster algorithms. For $p=2$, our running time is $O(\sum_{i=1}^q \text{nnz}(A_i) )$, which has no dependence on nnz$(b)$. For $p<2$, our running time is $O(\sum_{i=1}^q \text{nnz}(A_i) + \text{nnz}(b))$, which matches the prior best running time for $p=2$. We also consider the related all-pairs regression problem, where given $A \in \mathbb{R}^{n \times d}, b \in \mathbb{R}^n$, we want to solve $\min_{x} \|\bar{A}x - \bar{b}\|_p$, where $\bar{A} \in \mathbb{R}^{n^2 \times d}, \bar{b} \in \mathbb{R}^{n^2}$ consist of all pairwise differences of the rows of $A,b$. We give an $O(\text{nnz}(A))$ time algorithm for $p \in[1,2]$, improving the $\Omega(n^2)$ time needed to form $\bar{A}$. Finally, we initiate the study of Kronecker product low rank and low $t$-rank approximation. For input $\mathcal{A}$ as above, we give $O(\sum_{i=1}^q \text{nnz}(A_i))$ time algorithms, which is much faster than computing $\mathcal{A}$.
翻译:我们研究Kronecker产品回归问题, 其中设计矩阵是两个或两个以上基质的Kronecker产品。 鉴于美元=1,2,2美元,2美元=2美元,2美元=1美元,1美元=1美元,1美元=1美元,3美元=1美元,1美元=1美元,1美元=1美元,1美元=1美元,3美元=1美元,3美元=2美元=2美元,2美元=2美元=2美元,3美元=2美元=2美元,3美元=2美元=2美元,3美元=2美元=2美元=1美元,3美元=2美元=1美元=1美元,3美元=2美元=2美元=2美元, 美元=2美元=2美元=1美元, 美元=2美元=2美元=3美元, 美元=1美元=2美元, 美元=1美元=3美元 美元, 美元=1美元=1美元=2美元 美元 美元 美元 美元,这可以将我们(xxxxxxxx)