Abstract:Matrix sketching is aimed at finding compact representations of a matrix while simultaneously preserving most of its properties, which is a fundamental building block in modern scientific computing. Randomized algorithms represent state-of-the-art and have attracted huge interest from the fields of machine learning, data mining, and theoretic computer science. However, it still requires the use of the entire input matrix in producing desired factorizations, which can be a major computational and memory bottleneck in truly large problems. In this paper, we uncover an interesting theoretic connection between matrix low-rank decomposition and lossy signal compression, based on which a cascaded compression sampling framework is devised to approximate an m-by-n matrix in only O(m+n) time and space. Indeed, the proposed method accesses only a small number of matrix rows and columns, which significantly improves the memory footprint. Meanwhile, by sequentially teaming two rounds of approximation procedures and upgrading the sampling strategy from a uniform probability to more sophisticated, encoding-orientated sampling, significant algorithmic boosting is achieved to uncover more granular structures in the data. Empirical results on a wide spectrum of real-world, large-scale matrices show that by taking only linear time and space, the accuracy of our method rivals those state-of-the-art randomized algorithms consuming a quadratic, O(mn), amount of resources.

A Scalable CUR Matrix Decomposition Algorithm: Lower Time Complexity and Tighter Bound

Improving CUR Matrix Decomposition and the Nyström Approximation Via Adaptive Sampling

CUR Algorithm for Partially Observed Matrices

Efficient Spectrum-Revealing CUR Matrix Decomposition.

Low-rank approximation of parameter-dependent matrices via CUR decomposition

Randomized GCUR decompositions

A Simple Approach to Optimal CUR Decomposition

CUR from a Sparse Optimization Viewpoint

CURSOR: Scalable Mixed-Order Hypergraph Matching with CUR Decomposition

Towards more efficient SPSD matrix approximation and CUR matrix decomposition

A Restricted SVD type CUR Decomposition for Matrix Triplets

CUR and Generalized CUR Decompositions of Quaternion Matrices and their Applications

CPQR-based randomized algorithms for generalized CUR decompositions

More on low rank approximation of a matrix

Accuracy and Stability of CUR decompositions with Oversampling

CUR Low Rank Approximation of a Matrix at Sublinear Cost

Revisiting Co-Occurring Directions: Sharper Analysis and Efficient Algorithm for Sparse Matrices.

Column and row subset selection using nuclear scores: algorithms and theory for Nyström approximation, CUR decomposition, and graph Laplacian reduction

Randomization or Condensation?: Linear-Cost Matrix Sketching Via Cascaded Compression Sampling.