Options
How to extract a model from Robust PCA
jacobcybulski
Member, University Professor Posts: 391 Unicorn
in Help
In general, Robust PCA (rPCA) is an excellent algorithm capable of doing both PCA in the presence of outliers and identifying two classes of outliers, i.e. those that fall within the subset of principal components and those outside (which otherwise would not be detected after dimensionality reduction). Anomaly Detection extension offers rPCA, which implements the algorithm but it hides the PCA model. Is there a way of extracting the PCA model from rPCA? Or should rPCA be moved to the RM core with full features? (of course it is possible to do this in R or a hack in Python)
0
Best Answer
-
OptionsMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,510 RM Data ScientistHi @jacobcybulski ,as far as I know this is not possible. One would need to dig into the source code of rPCA in the extension to figure this one out.Cheers,Martin- Sr. Director Data Solutions, Altair RapidMiner -
Dortmund, Germany5
Answers
Dortmund, Germany