CoChat

Nonnegative/Binary matrix factorization with a D-Wave quantum annealer

Open access

The DOI matches its Crossref record; 55 of the 56 references it cites were found in a registry or an index.

Abstract

D-Wave quantum annealers represent a novel computational architecture and have attracted significant interest. Much of this interest has focused on the quantum behavior of D-Wave machines, and there have been few practical algorithms that use the D-Wave. Machine learning has been identified as an area where quantum annealing may be useful. Here, we show that the D-Wave 2X can be effectively used as part of an unsupervised machine learning method. This method takes a matrix as input and produces two low-rank matrices as output-one containing latent features in the data and another matrix describing how the features can be combined to approximately reproduce the input matrix. Despite the limited number of bits in the D-Wave hardware, this method is capable of handling a large input matrix. The D-Wave only limits the rank of the two output matrices. We apply this method to learn the features from a set of facial images and compare the performance of the D-Wave to two classical tools. This method is able to learn facial features and accurately reproduce the set of facial images. The performance of the D-Wave shows some promise, but has some limitations. It outperforms the two classical codes in a benchmark when only a short amount of computational time is allowed (200-20,000 microseconds), but these results suggest heuristics that would likely outperform the D-Wave in this benchmark.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0206653

Cites

55 of 55 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig 1 Face image reconstruction using features learned by NBMF. The five-by-seven matrix of images on the right shows the features that were learned. The two images on the left show the original image (top) and the reconstruction (bottom). The reconstruction is obtained by summing the features that are boxed in green. Note that although some of the features appear to be all black, they actually contain facial features that are small in magnitude (black corresponds to 0, white corresponds to 1).
Fig 1 Face image reconstruction using features learned by NBMF. The five-by-seven matrix of images on the right shows the features that were learned. The two images on the left show the original image (top) and the reconstruction (bottom). The reconstruction is obtained by summing the features that are boxed in green. Note that although some of the features appear to be all black, they actually contain facial features that are small in magnitude (black corresponds to 0, white corresponds to 1).CC0 · PubMed Central
Fig 2 The same features as shown in Fig 1 are shown. Here they are rescaled to maximize contrast so that the darkest pixel is black and the brightest pixel is white.
Fig 2 The same features as shown in Fig 1 are shown. Here they are rescaled to maximize contrast so that the darkest pixel is black and the brightest pixel is white.CC0 · PubMed Central
Fig 3 Cumulative time-to-targets for qbsolv (red) and Gurobi (blue) as a function of the number of annealing cycles executed by the D-Wave. When the number of anneals is 10, 100, or 1,000, the cumulative time-to-targets for both qbsolv and Gurobi exceeds the cumulative annealing time (orange). When the number of anneals is 10,000, the cumulative time-to-targets for qbsolv exceeds the cumulative annealing time, but Gurobi’s cumulative time-to-target is less than the cumulative annealing time. Note that in each of the 10, 100, and 1,000 anneal cases, 24,290 QUBOs were solved whereas only 19,432 QUBOs were solved in the 10,000 anneal cases. This was caused by earlier termination of the NBMF (algorithm 1).
Fig 3 Cumulative time-to-targets for qbsolv (red) and Gurobi (blue) as a function of the number of annealing cycles executed by the D-Wave. When the number of anneals is 10, 100, or 1,000, the cumulative time-to-targets for both qbsolv and Gurobi exceeds the cumulative annealing time (orange). When the number of anneals is 10,000, the cumulative time-to-targets for qbsolv exceeds the cumulative annealing time, but Gurobi’s cumulative time-to-target is less than the cumulative annealing time. Note that in each of the 10, 100, and 1,000 anneal cases, 24,290 QUBOs were solved whereas only 19,432 QUBOs were solved in the 10,000 anneal cases. This was caused by earlier termination of the NBMF (algorithm 1).CC0 · PubMed Central
Fig 4 The time-to-target for each instance of Eq 3 is shown for qbsolv (red dots), Gurobi (blue dots) in comparison to the annealing time used by the D-Wave (orange line). A different number of anneals was used in each execution of the NBMF (algorithm 1) ranging from 10 to 10,000.
Fig 4 The time-to-target for each instance of Eq 3 is shown for qbsolv (red dots), Gurobi (blue dots) in comparison to the annealing time used by the D-Wave (orange line). A different number of anneals was used in each execution of the NBMF (algorithm 1) ranging from 10 to 10,000.CC0 · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 55 references

Research works printed but not linked: 1

  1. Slawski M, Hein M, Lutsik P. Matrix factorization with binary components. In: Advances in Neural Information Processing Systems; 2013. p. 3210–3218.

Cited by

See all 149

Metadata

Authors
O’Malley, D., Vesselinov, V. V., Alexandrov, B. S., & Alexandrov, L. B.
Affiliations
Los Alamos National Laboratory (US)
Corresponding author
O’Malley, Daniel
Venue
PLOS ONE
Published
2018-12-10
Volume
13
Issue
12
Pages
e0206653
ISSN
1932-6203 (ISSN-L)
OpenAlex
W2605422054
Cited by
140
References
56
Citations by year
2026: 3 · 2025: 15 · 2024: 19 · 2023: 23 · 2022: 15 · 2021: 21 · 2020: 19 · 2019: 15 · 2018: 8
FWCI
8.75
Type
journal article
Language
en
Open access
Locations
  • PLoS ONE (publishedVersion, public-domain)
  • Europe PMC (PubMed Central) (publishedVersion, public-domain)
  • PubMed Central (publishedVersion, public-domain)
  • PubMed (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) (submittedVersion)
  • arXiv (Cornell University)
Funders
Los Alamos National Laboratory
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Quantum Computing Algorithms and Architecture
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Face and Expression Recognition
  • Physical Sciences › Computer Science › Artificial Intelligence › Stochastic Gradient Optimization Techniques
Keywords
D-Wave quantum annealer, non-negative matrix factorization, binary matrix factorization, quantum annealing, unsupervised feature learning, low-rank approximation, facial image features, quantum machine learning
MeSH terms
Machine Learning, Models, Theoretical, Quantum Theory

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Nonnegative/Binary matrix factorization with a D-Wave quantum annealer. https://app.cochat.ai/papers/10.1371/journal.pone.0206653
The paper
O’Malley, D., Vesselinov, V. V., Alexandrov, B. S., & Alexandrov, L. B. (2018). Nonnegative/Binary matrix factorization with a D-Wave quantum annealer. PLOS ONE. https://doi.org/10.1371/journal.pone.0206653

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.