PPLS: An efficient paralle algorithm for Partial Least-Squares
Resumo
PPLS, a parallel version for the Partial Least-Squares algorithm, is introduced in this article. The proposed algorithm is restricted to the case of only one dependent variable for the regression model. Unlike other approaches, such as the ones that calculate the PLS factors through Hebbian learning, PPLS is exact and does not depend on approximation or convergence criteria. In our experiments with a small data set, the algorithm shows a Speedup greater than 3 for the first 4 machines in a computer cluster architecture.
Palavras-chave:
Partial Least-Squares, PLS, parallelism, chemometrics, factor analysis
Referências
Michel Tenenhaus; Alain Morineau, editor. Les Méthodes PLS. Symposium International PLS'99. Cisia-Ceresta, October 1999.
Kenneth R. Beebe; Bruce R. Kowalski. An introduction to multivariate calibration and analysis. Analytical Chemistry, 59(17):1007–1017, September 1987.
Nicholas Carriero; David Gelernter. How to Write Parallel Programs: A First Course. MIT Press, Cambridge, 1991.
Ian Foster. Designing and Building Parallel Programs. Addison-Wesley, Cambridge, 1995.
Paul Geladi; Bruce R. Kowalski. Partial least squares regression: A tutorial. Analytica Chimica Acta, 185:1–17, 1986.
Fredric M. Ham; Ivica Kostanic. A neural network architecture for partial least-squares regression (plsnet) with supervised adaptive modular hebbian learning. Neural, Parallel & Scientific Computations, 6:35–72, 1998.
Fredric M. Ham; Thomas M. McDowall. Inverse model for simulation of partial least-squares regression: a robust neural network approach. In Steven K. Rogers, David B. Fogel, and James C. Bezdek (eds.), Applications and Science of Computational Intelligence, vol. 3390, pp. 36–47. SPIE, March 1998.
David M. Haaland; Edward V. Thomas. Partial least-squares methods for spectral analyses. 1. Relation to other quantitative calibration methods and the extraction of qualitative information. Analytical Chemistry, 60(11):1193–1202, June 1988.
David M. Haaland; Edward V. Thomas. Partial least-squares methods for spectral analyses. 2. Application to simulated and glass spectra data. Analytical Chemistry, 60(11):1202–1208, June 1988.
Santa Fe Institute. ftp://ftp.santafe.edu/pub/time-series/competition/, 1992.
Message Passing Interface. [link].
Joseph JáJá. An Introduction to Parallel Algorithms. Addison-Wesley Publishing Company, November 1992.
Ludovic Lebart; Alain Morineau; Marie Piron. Statistique Exploratoire Multidimensionnelle. Dunod, Paris, 1997.
Ruy Milidiú; Ricardo Machado; Raúl Rentería. Time series forecasting through wavelets and a mixture of experts models. Neurocomputing, 28:145–156, 1999.
E. Malthouse; R. Mah; A. Tamhane. Nonlinear partial least squares using neural networks. INCINC94 Chemometrics Conference, 1994.
Kenneth R. Beebe; Bruce R. Kowalski. An introduction to multivariate calibration and analysis. Analytical Chemistry, 59(17):1007–1017, September 1987.
Nicholas Carriero; David Gelernter. How to Write Parallel Programs: A First Course. MIT Press, Cambridge, 1991.
Ian Foster. Designing and Building Parallel Programs. Addison-Wesley, Cambridge, 1995.
Paul Geladi; Bruce R. Kowalski. Partial least squares regression: A tutorial. Analytica Chimica Acta, 185:1–17, 1986.
Fredric M. Ham; Ivica Kostanic. A neural network architecture for partial least-squares regression (plsnet) with supervised adaptive modular hebbian learning. Neural, Parallel & Scientific Computations, 6:35–72, 1998.
Fredric M. Ham; Thomas M. McDowall. Inverse model for simulation of partial least-squares regression: a robust neural network approach. In Steven K. Rogers, David B. Fogel, and James C. Bezdek (eds.), Applications and Science of Computational Intelligence, vol. 3390, pp. 36–47. SPIE, March 1998.
David M. Haaland; Edward V. Thomas. Partial least-squares methods for spectral analyses. 1. Relation to other quantitative calibration methods and the extraction of qualitative information. Analytical Chemistry, 60(11):1193–1202, June 1988.
David M. Haaland; Edward V. Thomas. Partial least-squares methods for spectral analyses. 2. Application to simulated and glass spectra data. Analytical Chemistry, 60(11):1202–1208, June 1988.
Santa Fe Institute. ftp://ftp.santafe.edu/pub/time-series/competition/, 1992.
Message Passing Interface. [link].
Joseph JáJá. An Introduction to Parallel Algorithms. Addison-Wesley Publishing Company, November 1992.
Ludovic Lebart; Alain Morineau; Marie Piron. Statistique Exploratoire Multidimensionnelle. Dunod, Paris, 1997.
Ruy Milidiú; Ricardo Machado; Raúl Rentería. Time series forecasting through wavelets and a mixture of experts models. Neurocomputing, 28:145–156, 1999.
E. Malthouse; R. Mah; A. Tamhane. Nonlinear partial least squares using neural networks. INCINC94 Chemometrics Conference, 1994.
Publicado
24/10/2000
Como Citar
MILIDIÚ, Ruy Luiz; RENTERÍA, Raúl.
PPLS: An efficient paralle algorithm for Partial Least-Squares. In: INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE AND HIGH PERFORMANCE COMPUTING (SBAC-PAD), 12. , 2000, São Pedro/SP.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2000
.
p. 359-363.
DOI: https://doi.org/10.5753/sbac-pad.2000.41235.
