Approximation Algorithms for $\ell_0$-Low Rank Approximation

We study the $\ell_0$-Low Rank Approximation Problem, where the goal is, given an $m \times n$ matrix $A$, to output a rank-$k$ matrix $A'$ for which $\|A'-A\|_0$ is minimized. Here, for a matrix $B$, $\|B\|_0$ denotes the number of its non-zero entries. This NP-hard variant of low rank approximation is natural for problems with no underlying metric, and its goal is to minimize the number of disagreeing data positions. We provide approximation algorithms which significantly improve the running time and approximation factor of previous work. For $k > 1$, we show how to find, in poly$(mn)$ time for every $k$, a rank $O(k \log(n/k))$ matrix $A'$ for which $\|A'-A\|_0 \leq O(k^2 \log(n/k)) \mathrm{OPT}$. To the best of our knowledge, this is the first algorithm with provable guarantees for the $\ell_0$-Low Rank Approximation Problem for $k > 1$, even for bicriteria algorithms. For the well-studied case when $k = 1$, we give a $(2+\epsilon)$-approximation in {\it sublinear time}, which is impossible for other variants of low rank approximation such as for the Frobenius norm. We strengthen this for the well-studied case of binary matrices to obtain a $(1+O(\psi))$-approximation in sublinear time, where $\psi = \mathrm{OPT}/\lVert A\rVert_0$. For small $\psi$, our approximation factor is $1+o(1)$.

[1]  Noga Alon,et al.  Deterministic Approximation Algorithms for the Nearest Codeword Problem , 2009, APPROX-RANDOM.

[2]  Arie Yeredor,et al.  Independent Component Analysis Over Galois Fields of Prime Order , 2011, IEEE Transactions on Information Theory.

[3]  Michael W. Mahoney Randomized Algorithms for Matrices and Data , 2011, Found. Trends Mach. Learn..

[4]  Fedor V. Fomin,et al.  Matrix Rigidity : Matrix Theory from the Viewpoint of Parameterized Complexity , 2016 .

[5]  Barnabás Póczos,et al.  Boolean Matrix Factorization and Noisy Completion via Message Passing , 2015, ICML.

[6]  Naren Ramakrishnan,et al.  Nonorthogonal decomposition of binary matrices for bounded-error data compression and analysis , 2006, TOMS.

[7]  Vijayalakshmi Atluri,et al.  The role mining problem: finding a minimal descriptive set of roles , 2007, SACMAT '07.

[8]  David P. Woodruff,et al.  Low rank approximation with entrywise l1-norm error , 2017, STOC.

[9]  Naren Ramakrishnan,et al.  Compression, clustering, and pattern discovery in very high-dimensional discrete-attribute data sets , 2005, IEEE Transactions on Knowledge and Data Engineering.

[10]  A. J. Walker New fast method for generating discrete random numbers with arbitrary frequency distributions , 1974 .

[11]  Heikki Mannila,et al.  A Simple Algorithm for Topic Identification in 0-1 Data , 2003, PKDD.

[12]  Chris H. Q. Ding,et al.  Binary Matrix Factorization with Applications , 2007, Seventh IEEE International Conference on Data Mining (ICDM 2007).

[13]  C. Papadimitriou,et al.  The complexity of massive data set computations , 2002 .

[14]  David P. Woodruff,et al.  Low rank approximation and regression in input sparsity time , 2013, STOC '13.

[15]  Marek Karpinski,et al.  Approximating minimum unsatisfiability of linear equations , 2002, SODA '02.

[16]  Silvio Lattanzi,et al.  Algorithms for $\ell_p$ Low-Rank Approximation , 2017, ICML.

[17]  Chris H. Q. Ding,et al.  Binary matrix factorization for analyzing gene expression data , 2009, Data Mining and Knowledge Discovery.

[18]  David P. Woodruff,et al.  Weighted low rank approximations with provable guarantees , 2016, STOC.

[19]  Kristoffer Arnsfelt Hansen,et al.  Low Rank Approximation of Binary Matrices: Column Subset Selection and Generalizations , 2015, MFCS.

[20]  Jieping Ye,et al.  Mining discrete patterns via binary matrix factorization , 2009, KDD.

[21]  B. Schölkopf,et al.  Modeling Dyadic Data with Binary Latent Factors , 2007 .

[22]  David P. Woodruff,et al.  Input Sparsity and Hardness for Robust Subspace Approximation , 2015, 2015 IEEE 56th Annual Symposium on Foundations of Computer Science.

[23]  Nicolas Gillis,et al.  On the Complexity of Robust PCA and ℓ1-norm Low-Rank Matrix Approximation , 2015, Math. Oper. Res..

[24]  Milos Hauskrecht,et al.  Noisy-OR Component Analysis and its Application to Link Analysis , 2006, J. Mach. Learn. Res..

[25]  Michael Alekhnovich More on Average Case vs Approximation Complexity , 2011, computational complexity.

[26]  Pauli Miettinen,et al.  The Discrete Basis Problem , 2006, IEEE Transactions on Knowledge and Data Engineering.

[27]  Arie Yeredor,et al.  ICA over finite fields - Separability and algorithms , 2012, Signal Process..

[28]  Yi Ma,et al.  Robust principal component analysis? , 2009, JACM.

[29]  Vilém Vychodil,et al.  Discovery of optimal factors in binary data via a novel method of matrix decomposition , 2010, J. Comput. Syst. Sci..

[30]  Huy L. Nguyen,et al.  OSNAP: Faster Numerical Linear Algebra Algorithms via Sparser Subspace Embeddings , 2012, 2013 IEEE 54th Annual Symposium on Foundations of Computer Science.

[31]  David P. Woodruff Sketching as a Tool for Numerical Linear Algebra , 2014, Found. Trends Theor. Comput. Sci..

[32]  Richard M. Karp,et al.  An optimal algorithm for Monte Carlo estimation , 1995, Proceedings of IEEE 36th Annual Foundations of Computer Science.

[33]  Tao Li,et al.  A general model for clustering binary data , 2005, KDD '05.

[34]  Michael W. Mahoney,et al.  Low-distortion subspace embeddings in input-sparsity time and applications to robust linear regression , 2012, STOC '13.

[35]  Sanjeev Arora,et al.  Computing a nonnegative matrix factorization -- provably , 2011, STOC '12.

[36]  Peng Jiang,et al.  A Clustering Approach to Constrained Binary Matrix Factorization , 2014 .

[37]  David P. Woodru Sketching as a Tool for Numerical Linear Algebra , 2014 .

[38]  Ananth Grama,et al.  PROXIMUS: a framework for analyzing very high dimensional discrete-attributed datasets , 2003, KDD '03.

[39]  Shang-Hua Teng,et al.  Smoothed Analysis of the Condition Numbers and Growth Factors of Matrices , 2003, SIAM J. Matrix Anal. Appl..

[40]  Leslie G. Valiant,et al.  Graph-Theoretic Arguments in Low-Level Complexity , 1977, MFCS.

[41]  Saharon Rosset,et al.  Generalized Independent Component Analysis Over Finite Alphabets , 2016, IEEE Trans. Inf. Theory.

[42]  Pauli Miettinen,et al.  MDL4BMF: Minimum Description Length for Boolean Matrix Factorization , 2014, TKDD.