Abstract
In this work we present the empirical influence functions for the covariances (eigenvalues) and directions (eigenvectors) of partial least squares under the constraint of uncorrelated components. We apply the results to several data sets and provide advice for using these tools in practice.
| Original language | English |
|---|---|
| Pages (from-to) | 293-306 |
| Number of pages | 14 |
| Journal | Computational Statistics |
| Volume | 22 |
| Issue number | 2 |
| DOIs | |
| State | Published - Jul 2007 |
Keywords
- Influence function analysis
- Outliers
- Partial least squares
ASJC Scopus subject areas
- Statistics and Probability
- Statistics, Probability and Uncertainty
- Computational Mathematics