What is the exact difference between principal component analysis (PCA)and principal component regression (PCR)? With what kind of data sets can one use these methods?
Asked
Active
Viewed 6,616 times
2
-
8See Is principal component regression (PCR) using principal component scores for regression? & How to use principal components as predictors in GLM?. Examples of PCA where PCs with low variance are “useful” is also interesting. – Scortchi - Reinstate Monica Dec 01 '15 at 12:55
-
The question does not show much research effort, but it is formulated such that it can be quite useful for google searches in the future. I don't agree with the downvote. +1. – amoeba Dec 01 '15 at 14:32
-
@Fati, if you think that Peter answered your question, please "accept" his answer by pressing on a green tick to the left of it. – amoeba Dec 15 '15 at 21:15
1 Answers
5
Principal component analysis is a method of data reduction - representing a large number of variable by a (much) smaller number, each of which is a linear combination of the original variables.
One output of PCA is principal component scores. Principal component regression uses those scores as independent variables in a regression.
Peter Flom
- 119,535
- 36
- 175
- 383
-
is it possible explain me with matlab code or with example for example if i give one data set with these method what is different in out put of these methods ? – Fati Dec 01 '15 at 13:18
-
2@fati See references in comment from Scortchi above. These methods are well documented here and elsewhere. Please focus on asking questions that are new here; otherwise the forum is a resource for your study. – Nick Cox Dec 01 '15 at 13:28
-
-
I don't know Matlab but the output is completely different. PCA is a necessary step towards PC regression. But you can certainly do PCA without ever doing a regression. – Peter Flom Dec 01 '15 at 23:49