Dear MachineLearning Task View maintainers,
I would like to suggest adding the CRAN package msPCA to the Machine Learning & Statistical Learning task view.
Package: msPCA
CRAN: https://cran.r-project.org/package=msPCA
Documentation: https://jeanpauphilet.github.io/msPCA/
Reference: Cory-Wright and Pauphilet, “Sparse PCA with Multiple Components,” Operations Research, 2026, doi:10.1287/opre.2023.0598
msPCA implements sparse principal component analysis with multiple components. It provides a regularized unsupervised dimension-reduction method for high-dimensional data, producing sparse loading vectors while explicitly controlling redundancy across components. The package supports either orthogonality of loading vectors or zero pairwise correlation of component scores.
I think it fits this task view as a method for unsupervised learning, sparse representation, and dimensionality reduction. I could not find an 'Unsupervised learning/clustering' section, which would have been the most obvious fit. Alternatively, it could fit in the 'Regularized and Shrinkage Methods' or 'Other procedures' sections.
A possible short entry could be:
“The msPCA package implements sparse principal component analysis with multiple components, providing regularized unsupervised dimension reduction for high-dimensional data.”
Thank you for considering it.
Dear MachineLearning Task View maintainers,
I would like to suggest adding the CRAN package
msPCAto the Machine Learning & Statistical Learning task view.Package:
msPCACRAN: https://cran.r-project.org/package=msPCA
Documentation: https://jeanpauphilet.github.io/msPCA/
Reference: Cory-Wright and Pauphilet, “Sparse PCA with Multiple Components,” Operations Research, 2026, doi:10.1287/opre.2023.0598
msPCAimplements sparse principal component analysis with multiple components. It provides a regularized unsupervised dimension-reduction method for high-dimensional data, producing sparse loading vectors while explicitly controlling redundancy across components. The package supports either orthogonality of loading vectors or zero pairwise correlation of component scores.I think it fits this task view as a method for unsupervised learning, sparse representation, and dimensionality reduction. I could not find an 'Unsupervised learning/clustering' section, which would have been the most obvious fit. Alternatively, it could fit in the 'Regularized and Shrinkage Methods' or 'Other procedures' sections.
A possible short entry could be:
“The
msPCApackage implements sparse principal component analysis with multiple components, providing regularized unsupervised dimension reduction for high-dimensional data.”Thank you for considering it.