Advertisement

Matrix Completion Under Interval Uncertainty: Highlights

  • Jakub MarecekEmail author
  • Peter Richtarik
  • Martin Takac
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11053)

Abstract

We present an overview of inequality-constrained matrix completion, with a particular focus on alternating least-squares (ALS) methods. The simple and seemingly obvious addition of inequality constraints to matrix completion seems to improve the statistical performance of matrix completion in a number of applications, such as collaborative filtering under interval uncertainty, robust statistics, event detection, and background modelling in computer vision. An ALS algorithm MACO by Marecek et al. outperforms others, including Sparkler, the implementation of Li et al. Code related to this paper is available at: http://optml.github.io/ac-dc/.

Notes

Acknowledgement

The work of JM received funding from the European Union’s Horizon 2020 Programme (Horizon2020/2014-2020) under grant agreement No. 688380. The work of MT was partially supported by the U.S. National Science Foundation, under award numbers NSF:CCF:1618717, NSF:CMMI:1663256, and NSF:CCF:1740796. PR acknowledges support from KAUST Faculty Baseline Research Funding Program.

References

  1. 1.
    Akhriev, A., Marecek, J., Simonetto, A.: Pursuit of low-rank models of time-varying matrices robust to sparse and measurement noise. Preprint arXiv:1809.03550 (2018, submitted)
  2. 2.
    Dutta, A., Li, X., Richtarik, P.: Weighted low-rank approximation of matrices and background modeling. Preprint arXiv:1804.06252 (2018, submitted)
  3. 3.
    Jahrer, M., Töscher, A., Legenstein, R.: Combining predictions for accurate recommender systems. In: KDD, pp. 693–702. ACM (2010)Google Scholar
  4. 4.
    Li, B., Tata, S., Sismanis, Y.: Sparkler: supporting large-scale matrix factorization. In: EDBT, pp. 625–636. ACM (2013)Google Scholar
  5. 5.
    Li, X., Dutta, A.: Weighted low rank approximation for background estimation problems. In: ICCVW, pp. 1853–1861, October 2017Google Scholar
  6. 6.
    Li, X., Dutta, A., Richtarik, P.: A batch-incremental video background estimation model using weighted low-rank approximation of matrices. In: ICCVW, pp. 1835–1843, October 2017Google Scholar
  7. 7.
    Marecek, J., Maroulis, S., Kalogeraki, V., Gunopulos, D.: Low-rank methods in event detection. Preprint arXiv:1802.03649 (2018, submitted)
  8. 8.
    Marecek, J., Richtarik, P., Takac, M.: Matrix completion under interval uncertainty. Eur. J. Oper. Res. 256(1), 35–43 (2017)MathSciNetCrossRefGoogle Scholar
  9. 9.
    Marecek, J., Simonetto, A., Maroulis, S., Kalogeraki, V., Gunopulos, D.: Low-rank subspace pursuit in event detection (2018, submitted)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.IBM Research – IrelandDublin 15Ireland
  2. 2.School of MathematicsUniversity of EdinburghEdinburghUK
  3. 3.KAUSTThuwalKingdom of Saudi Arabia
  4. 4.Department of Industrial and Systems EngineeringLehigh UniversityBethlehemUSA

Personalised recommendations