I want to perform dimensionality reduction (in particular, PCA) on a data set that is highly seasonal. One approach that I came across when researching this is "seasonal PCA", where you split your data set into seasons, and then run PCA on each season separately. This seems like it wastes a lot of data.

Are there other approaches for dealing with seasonal variations when doing PCA, or is there perhaps another dimensionality reduction approach that's more appropriate?

  • $\begingroup$ Can you be more specific about what you want to achieve with PCA? Why are you doing it, what do you want to get out of it? How exactly is a "seasonality" a problem for you? $\endgroup$ – amoeba Jan 21 '15 at 15:34
  • 1
    $\begingroup$ I'm using it to model a group of financial products that have seasonality in their prices. Lots of financial products are like this, e.g. energy, crops, etc. The goal is to use it to determine risk exposure. Seasonality is a problem because the group of products will have different principle components during one season versus another. $\endgroup$ – Thomas Johnson Jan 21 '15 at 15:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.