Randomized Nonnegative Matrix Factorization

N. Benjamin Erichson, Ariana Mendible, Sophie Wihlborn, J. Nathan Kutz
11/6/2017
stat.MLcs.CV

Official code: erichson/ristretto (118 stars, GPL-3.0, last commit 2021-04-17).

Abstract

Nonnegative matrix factorization (NMF) is a powerful tool for data mining. However, the emergence of `big data' has severely challenged our ability to compute this fundamental decomposition using deterministic algorithms. This paper presents a randomized hierarchical alternating least squares (HALS) algorithm to compute the NMF. By deriving a smaller matrix from the nonnegative input data, a more efficient nonnegative decomposition can be computed. Our algorithm scales to big data applications while attaining a near-optimal factorization. The proposed algorithm is evaluated using synthetic and real world data and shows substantial speedups compared to deterministic HALS.

View on arXivView PDF

Code Implementations(1)

erichson/ristrettoOfficial100%
GPL-3.0

Randomized Dimension Reduction Library

11825Python, Jupyter NotebookMay 18, 20175 years agoGPL-3.0
dimension-reductiondynamic-mode-decompositionlu-decompositionmatrix-approximationnonnegative-matrix-factorization+6 more

Similar Papers

RSVDPACK: An implementation of randomized algorithms for computing the singular value, interpolative, and CUR decompositions of matrices on multi-core and GPU architectures

81% match1 repo1052015

Nonnegative Matrix Factorization in the Component-Wise L1 Norm for Sparse Data

80% match2026

Learning nonnegative matrix factorizations from compressed data

80% match2024

Randomized Matrix Decompositions using R

80% match1 repo1052016

Algorithms for Boolean Matrix Factorization using Integer Programming

79% match2023

Majorization-minimization for Sparse Nonnegative Matrix Factorization with the β-divergence

79% match2022

Cite this paper

@article{erichson2017randomized,
  title  = {Randomized Nonnegative Matrix Factorization},
  author = {N. Benjamin Erichson and Ariana Mendible and Sophie Wihlborn and J. Nathan Kutz},
  year   = {2017},
  eprint = {1711.02037},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1711.02037}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1711.02037.svg)](https://codeofpaper.com/paper/1711.02037)

Discussion