Skip to Main content Skip to Navigation
New interface
Journal articles

Data-Driven Sparse Partial Least Squares

Hadrien Lorenzo 1 Olivier Cloarec 2 Rodolphe Thiébaut 3, 4, 5 Jérôme Saracco 1 
1 ASTRAL - Méthodes avancées d’apprentissage statistique et de contrôle
IMB - Institut de Mathématiques de Bordeaux, Inria Bordeaux - Sud-Ouest, UB - Université de Bordeaux, Bordeaux INP - Institut Polytechnique de Bordeaux, Naval Group
3 SISTM - Statistics In System biology and Translational Medicine
Inria Bordeaux - Sud-Ouest, BPH - Bordeaux population health
Abstract : In the supervised high dimensional settings with a large number of variables and a low number of individuals, variable selection allows a simpler interpretation and more reliable predictions. That subspace selection is often managed with supervised tools when the real question is motivated by variable prediction. We propose a Partial Least Square (PLS) based method, called data-driven sparse PLS (ddsPLS), allowing variable selection both in the covariate and the response parts using a single hyper-parameter per component. The subspace estimation is also performed by tuning a number of underlying parameters. The ddsPLS method is compared to existing methods such as classical PLS and two well established sparse PLS methods through numerical simulations. The observed results are promising both in terms of variable selection and prediction performance. This methodology is based on new prediction quality descriptors associated with the classical R 2 and Q 2 and uses bootstrap sampling to tune parameters and select an optimal regression model.
Document type :
Journal articles
Complete list of metadata

https://hal.inria.fr/hal-03368956
Contributor : Hadrien Lorenzo Connect in order to contact the contributor
Submitted on : Thursday, October 7, 2021 - 9:49:09 AM
Last modification on : Wednesday, October 26, 2022 - 4:05:37 AM
Long-term archiving on: : Saturday, January 8, 2022 - 6:19:58 PM

File

ddsPLS_tout_seul_no_bib_Pour_H...
Files produced by the author(s)

Identifiers

Citation

Hadrien Lorenzo, Olivier Cloarec, Rodolphe Thiébaut, Jérôme Saracco. Data-Driven Sparse Partial Least Squares. Statistical Analysis and Data Mining, 2021, ⟨10.1002/sam.11558⟩. ⟨hal-03368956⟩

Share

Metrics

Record views

69

Files downloads

131