can you please explain in principal component analysis (PCA), why maximum variance means minimum information loss?
Asked
Active
Viewed 69 times
0
-
1It doesn't except for certain purposes, with information measured in certain ways. (A standard counterexample is principal components regression, where the relevant information might, in the extreme, all reside in the component with least variance.) You therefore might want to provide a little more context for your question. – whuber Apr 20 '23 at 22:25
-
do you know why we want to maximize variance when we are computing the PCA? Is maximizing variance the same as retaining as much information as possible? thanks – shawn Apr 21 '23 at 00:33