We use Principal Component Analysis for dimension reduction, why is it important?
Asked
Active
Viewed 34 times
1
-
1To deal with `Curse of Dimensionality`. – Sociopath Jul 31 '18 at 09:45
-
1To reduce dimensionality. – Eran Moshe Jul 31 '18 at 11:36
-
This is two-fold question, actually two questions. (1) Why would we use dim. red. techniques like PCA, which do not belong to classification domain, _before_ a classification. Is it useful? (2) Why some classification techniques, like LDA (discrim. analysis), are themselves special dim. reduction methods; what's the reason to do dim. red. _on the way_ of classifying? – ttnphns Jul 31 '18 at 12:44