Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
1145742 | Journal of Multivariate Analysis | 2014 | 13 Pages |
Abstract
For multiple multivariate datasets, we derive conditions under which Generalized Canonical Correlation Analysis improves classification performance of the projected datasets, compared to standard Canonical Correlation Analysis using only two data sets. We illustrate our theoretical results with simulations and a real data experiment.
Keywords
Related Topics
Physical Sciences and Engineering
Mathematics
Numerical Analysis
Authors
Cencheng Shen, Ming Sun, Minh Tang, Carey E. Priebe,