On the Problems of Construction and Statistical Inference Associated with a Generalization of Canonical Variables.
Abstract
Canonical correlation analysis provides a means by which the relation between two sets of variates can be assessed. The present study is concerned with an extension to K sets of variates. As in the case of two sets of variates, linear combinations of the variates are used, but now alternative criteria need be defined. The criteria developed are based on the generalized variance. Properties and tests using this criterion are obtained. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1982
- Accession Number
- ADA115315
Entities
People
- Ashis Sen Gupta
Organizations
- Stanford University