Big data in multi-block data analysis: An approach to parallelizing Partial Least Squares Mode B algorithm
Partial Least Squares (PLS) Mode B is a multi-block method and a tightly coupled algorithm for estimating structural equation models (SEMs). Describing key aspects of parallel computing, we approach the parallelization of the PLS Mode B algorithm to operate on large distributed data. We show the sca...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2019-04-01
|
| Series: | Heliyon |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2405844018367616 |
| _version_ | 1818843287436394496 |
|---|---|
| author | Alba Martinez-Ruiz Cristina Montañola-Sales |
| author_facet | Alba Martinez-Ruiz Cristina Montañola-Sales |
| author_sort | Alba Martinez-Ruiz |
| collection | DOAJ |
| description | Partial Least Squares (PLS) Mode B is a multi-block method and a tightly coupled algorithm for estimating structural equation models (SEMs). Describing key aspects of parallel computing, we approach the parallelization of the PLS Mode B algorithm to operate on large distributed data. We show the scalability and performance of the algorithm at a very fine-grained level thanks to the versatility of pbdR, a R-project library for parallel computing. We vary several factors under different data distribution schemes in a supercomputing environment. Shorter elapsed times are obtained for the square-blocking factor 16×16 using a grid of processors as square as possible and non-square blocking factors 1000×4 and 10000×4 using an one-column grid of processors. Depending on the configuration, distributing data in a larger number of cores allows reaching speedups of up to 121 over the CPU implementation. Moreover, we show that SEMs can be estimated with big data sets using current state-of-the-art algorithms for multi-block data analysis. |
| first_indexed | 2024-12-19T04:55:28Z |
| format | Article |
| id | doaj.art-e2907823e2bf423faf8fcaaea4035a49 |
| institution | Directory Open Access Journal |
| issn | 2405-8440 |
| language | English |
| last_indexed | 2024-12-19T04:55:28Z |
| publishDate | 2019-04-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Heliyon |
| spelling | doaj.art-e2907823e2bf423faf8fcaaea4035a492022-12-21T20:35:15ZengElsevierHeliyon2405-84402019-04-0154e01451Big data in multi-block data analysis: An approach to parallelizing Partial Least Squares Mode B algorithmAlba Martinez-Ruiz0Cristina Montañola-Sales1Universidad Católica de la Santísima Concepción, Alonso de Ribera 2850, Concepción, Chile; Corresponding author.IQS-Universitat Ramon Llull (URL), Via Augusta, 390, 08017 Barcelona, Spain; Barcelona Supercomputing Center, Centro Nacional de Supercomputación (BSC-CNS), Jordi Girona 29, 08034, Barcelona, SpainPartial Least Squares (PLS) Mode B is a multi-block method and a tightly coupled algorithm for estimating structural equation models (SEMs). Describing key aspects of parallel computing, we approach the parallelization of the PLS Mode B algorithm to operate on large distributed data. We show the scalability and performance of the algorithm at a very fine-grained level thanks to the versatility of pbdR, a R-project library for parallel computing. We vary several factors under different data distribution schemes in a supercomputing environment. Shorter elapsed times are obtained for the square-blocking factor 16×16 using a grid of processors as square as possible and non-square blocking factors 1000×4 and 10000×4 using an one-column grid of processors. Depending on the configuration, distributing data in a larger number of cores allows reaching speedups of up to 121 over the CPU implementation. Moreover, we show that SEMs can be estimated with big data sets using current state-of-the-art algorithms for multi-block data analysis.http://www.sciencedirect.com/science/article/pii/S2405844018367616Computer scienceComputational mathematics |
| spellingShingle | Alba Martinez-Ruiz Cristina Montañola-Sales Big data in multi-block data analysis: An approach to parallelizing Partial Least Squares Mode B algorithm Heliyon Computer science Computational mathematics |
| title | Big data in multi-block data analysis: An approach to parallelizing Partial Least Squares Mode B algorithm |
| title_full | Big data in multi-block data analysis: An approach to parallelizing Partial Least Squares Mode B algorithm |
| title_fullStr | Big data in multi-block data analysis: An approach to parallelizing Partial Least Squares Mode B algorithm |
| title_full_unstemmed | Big data in multi-block data analysis: An approach to parallelizing Partial Least Squares Mode B algorithm |
| title_short | Big data in multi-block data analysis: An approach to parallelizing Partial Least Squares Mode B algorithm |
| title_sort | big data in multi block data analysis an approach to parallelizing partial least squares mode b algorithm |
| topic | Computer science Computational mathematics |
| url | http://www.sciencedirect.com/science/article/pii/S2405844018367616 |
| work_keys_str_mv | AT albamartinezruiz bigdatainmultiblockdataanalysisanapproachtoparallelizingpartialleastsquaresmodebalgorithm AT cristinamontanolasales bigdatainmultiblockdataanalysisanapproachtoparallelizingpartialleastsquaresmodebalgorithm |