sgcrfpy icon indicating copy to clipboard operation
sgcrfpy copied to clipboard

compare performance between covariance and correlation matrices

Open dswah opened this issue 9 years ago • 1 comments

calvin mentioned that in some cases its better to use the correlation matrix...

dswah avatar Aug 08 '16 21:08 dswah

i think this is about normalizing the data so that each variable has unit variance.

this helps us compare variables that have different scales.

dswah avatar Oct 17 '16 21:10 dswah