CI-EIG
A procedure for calculating confidence intervals for the eigendecomposition of covariance matrices.
Description:
We provide a procedure to obtain distribution-free confidence intervals on the eigendecomposition of covariance matrices by using U-statistics, Weyl’s theorem and the Eigenvalue-eigenvector identity.
More Information:
Originates from the AI-Driven Data Science research challenge (FAIR 2019–2023).
Responsible Research Lead: Matthew B. Blaschko
Authors: Teodora Popordanoska, Aleksei Tiulpin, Wacha Bounliphone, Matthew B. Blaschko
Programming Language: Python
Categories: Unsupervised Learning
Publications:
T. Popordanoska, A. Tiulpin, W. Bounliphone M. Blaschko, "Distribution-independent confidence intervals for the eigendecomposition of covariance matrices via the eigenvalue-eigenvector identity", ICML 2021 workshop on distribution-free uncertainty quantification, 2021