Data prediction based on a PCA model

8 views (last 30 days)
toka55
toka55 on 4 Dec 2017
Answered: Elizabeth Reese on 6 Dec 2017
I try to compare 5 data sets (1 training, 4 test sets) based on a PCA model calclulated from a training data set. The idea is to calculate PCA of data set 1 and then apply the data sets 2-5 to PCA model from data 1. Finally, I would like to overlay all data sets in a score diagram to show the differences.
Is there a function available in Matlab for this purpose?

Answers (1)

Elizabeth Reese
Elizabeth Reese on 6 Dec 2017
There is the pca function and svd function in MATLAB that you can use to do this. There is also this Answers post which explains how to take a new point and project it to the PCA basis. You will have to calculate the scores over again for the other data sets.

Tags

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!