Conference Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 156 - Contributed Poster Presentations: Section on Statistics in Genomics and Genetics
Type: Contributed
Date/Time: Monday, August 8, 2022 : 10:30 AM to 12:20 PM
Sponsor: Section on Statistics in Genomics and Genetics
Abstract #323211
Title: Expanding Sparse Partial Least Squares Regression Using Dynamic Bootstrap
Author(s): Frédéric Bertrand* and Myriam Maumy
Companies: Troyes Technology University and Troyes Technology University
Keywords: partial least squares regression; sparse partial least squares regression; hyperparameters; bootstrap; cross validation; stability
Abstract:

Methods based on partial least squares (PLS) regression, which has recently gained much attention in the analysis of high-dimensional single or multi-omics datasets, have been developed since the early 2000s for performing variable selection. Most of these techniques rely on tuning parameters often determined by cross-validation (CV) based methods, which raises essential stability issues. We have developed a new dynamic bootstrap-based method for significant predictor selection, suitable for both PLS regression and its incorporation into generalized linear models (GPLS). It relies on establishing bootstrap confidence intervals, which allows testing of the significance of predictors at preset type I risk ?, and avoids CV. We have also developed adapted versions of sparse PLS and sparse GPLS regression, using a recently introduced non-parametric bootstrap-based technique to determine the numbers of components. We compare their variable selection reliability and stability concerning tuning parameters determination and their predictive ability, using simulated data for PLS and real microarray gene expression data for PLS-logistic classification. Implemented in the bootPLS R package.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2022 program