I am trying to understand how CCA can be formulated as a least-squares problem in the binary class case.
I understood how CCA and Least square problem works. But not getting how we can formulate a CCA to solve least square problem.
$\begingroup$
$\endgroup$
2
-
1$\begingroup$ Well, it's been a while, but have you read "A Least Squares Formulation for Canonical Correlation Analysis" citeseerx.ist.psu.edu/viewdoc/… $\endgroup$– Mark L. StoneCommented Apr 29, 2018 at 17:50
-
$\begingroup$ My answer to this question may be helpful to you. stats.stackexchange.com/questions/309031/… $\endgroup$– eric_kernfeldCommented Sep 18, 2018 at 4:20
Add a comment
|