Abstract
Data reduction algorithms often produce inaccurate results for loss of relevant information. Recently, the singular value decomposition (SVD) method has been used as preprocessing method in order to deal with high-dimensional data and achieve fuzzy-rough reduct convergence on higher dimensional datasets. Despite the well-known fact that SVD offers attractive properties, its high computational cost remains a critical issue. In this work, we present a parallel implementation of the SVD algorithm on graphics processing units using CUDA programming model. Our approach is based on an iterative parallel version of the QR factorization by means of Givens rotations using the Sameh and Kuck scheme. Our results show significant improvements in terms of performances with respect to the CPU version that encourage its usability for this expensive processing of data.
Similar content being viewed by others
References
Andrews H, Patterson C (1976) Singular value decompositions and digital image processing. IEEE Trans Acoust Speech Signal Process 24(1):26–53
Cuomo S, Galletti A, Giunta G, Marcellino L (2015) Toward a multi-level parallel framework on GPU cluster with petsc-cuda for pde-based optical flow computation. Proc Comput Sci 51(1):170–179. doi:10.1016/j.procs.2015.05.220
Cuomo S, Galletti A, Marcellino L (2015) A GPU algorithm in a distributed computing system for 3d MRI denoising. In: Proceedings—2015 10th international conference on P2P, parallel, grid, cloud and internet computing, 3PGCIC 2015, pp 557–562
Cuomo S, De Michele P, Galletti A, Marcellino L (2016) A GPU parallel implementation of the local principal component analysis overcomplete method for DW image denoising. In: Proceedings—IEEE symposium on computers and communications. pp 26-31. doi:10.1109/ISCC.2016.7543709
Cuomo S, De Michele P, Galletti A, Marcellino L (2016) A parallel PDE-based numerical algorithm for computing the optical flow in hybrid systems. J Comput Sci. doi:10.1016/j.jocs.2017.03.011. Article in Press
Cuomo S, De Michele P, Galletti A, Marcellino L (2016) Local principal component analysis overcomplete method: a GPU parallel implementation combining shared and global memories. In: International conference on high performance computing and simulation, HPCS, pp 81–87. doi:10.1109/HPCSim.2016.7568319
Cuomo S, De Michele P, Maiorano F, Marcellino L (2016) Gpu profiling of singular value decomposition in olpca method for image denoising. In: International conference on P2P. Cloud and internet computing. Springer, Parallel, Grid, pp 707–716
D’Amore L, Marcellino L, Mele V, Romano D (2012) Deconvolution of 3D fluorescence microscopy images using graphics processing units. In: Lecture notes in computer science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes. Bioinformatics vol. 7203, no. 1, pp 690–699
Golub RC, Reinsch C (1970) Singular value decomposition and least squares solutions. Numer Math 14:403–420
Jensen R, Shen Q (2005) Fuzzy-rough data reduction with ant colony optimization. Fuzzy Sets Syst 149(1):5–20
Pudil P, Novoviov J (1998) Novel methods for feature subset selection with respect to problem knowledge. In: Liu H, Motoda H (eds). Feature extraction, construction and selection. p. 101. doi:10.1007/978-1-4615-5725-8-7. ISBN 978-1-4613-7622-4
Quafafou M, Boussouf M (2000) Generalized rough sets based feature selection. J Intell Data Anal 4(1):3–17
Rama Devi Y, Venu Gopal P, Sai Prasad P (2011) Fuzzy rough data reduction using SVD. Int J Comput Electr Eng 3(3):384–388
Richard J, Shen Q (2002) Fuzzy-rough sets for descriptive dimensionality reduction. In: Fuzzy systems, proceedings of the 2002 IEEE international conference, pp 29–34
Roweis ST, Saul LK (2000) Nonlinear dimensionality reduction by locally linear embedding. Science 290(5500):2323–2326. doi:10.1126/science.290.5500.2323
Sameh KD, Kuck AH (1978) On stable parallel linear system solvers. J ACM 25(1):81–91
Shen Q, Chouchoulas A (2000) A modular approach to generating fuzzy rules with reduced attributes for the monitoring of complex systems. Eng Appl Artif Intell 13(3):263–278
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Funding
This paper is not funded by a project.
Conflict of interest
All authors declare the absence of conflicts of interest.
Ethical approval
This article does not contain any studies with human participants or animals performed by any of the authors.
Additional information
Communicated by M. Anisetti.
Rights and permissions
About this article
Cite this article
Cuomo, S., Galletti, A., Marcellino, L. et al. On GPU–CUDA as preprocessing of fuzzy-rough data reduction by means of singular value decomposition. Soft Comput 22, 1525–1532 (2018). https://doi.org/10.1007/s00500-017-2887-x
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00500-017-2887-x