亚线性时间正定矩阵的低秩近似
Sublinear Time Low-Rank Approximation of Positive Semidefinite Matrices
AI总结:
本文提出了一种在亚线性时间内计算正定矩阵低秩近似的算法,通过因子形式输出一个秩为k的矩阵B,使得B与原矩阵A的F范数平方误差不超过(1+ε)倍的最优秩k近似A_k的F范数平方误差,并在特定条件下无需读取全部矩阵元素。
摘要:
我们展示了如何在亚线性时间内计算任意正定矩阵的相对误差低秩近似。即对于任意n×n的正定矩阵A,在~O(n·poly(k/ε))时间内输出一个因子形式的秩k矩阵B,使得||A-B||_F^2 ≤ (1+ε)||A-A_k||_F^2,其中A_k是A的最佳秩k近似。当k和1/ε与A的稀疏性相比不太大时,我们的算法不需要读取矩阵的所有元素。因此,我们显著改进了基于无偏子空间嵌入的先前O(nnz(A))时间算法,并绕过了通用矩阵的O(nnz(A))时间下界。我们证明了正定矩阵低秩近似的时界,显示我们的算法接近最优。最后,我们扩展了我们的技术,以在(通常更强的)谱范数度量||A-B||_2^2下给出低秩近似的亚线性时间算法,并在正定矩阵上进行岭回归。
英文摘要:
We show how to compute a relative-error low-rank approximation to any positive semidefinite (PSD) matrix in sublinear time, i.e., for any $n \times n$ PSD matrix $A$, in $\tilde O(n \cdot poly(k/ε))$ time we output a rank-$k$ matrix $B$, in factored form, for which $\|A-B\|_F^2 \leq (1+ε)\|A-A_k\|_F^2$, where $A_k$ is the best rank-$k$ approximation to $A$. When $k$ and $1/ε$ are not too large compared to the sparsity of $A$, our algorithm does not need to read all entries of the matrix. Hence, we significantly improve upon previous $nnz(A)$ time algorithms based on oblivious subspace embeddings, and bypass an $nnz(A)$ time lower bound for general matrices (where $nnz(A)$ denotes the number of non-zero entries in the matrix). We prove time lower bounds for low-rank approximation of PSD matrices, showing that our algorithm is close to optimal. Finally, we extend our techniques to give sublinear time algorithms for low-rank approximation of $A$ in the (often stronger) spectral norm metric $\|A-B\|_2^2$ and for ridge regression on PSD matrices.
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。