PLS-DA prediction when the sample class are unknown

Hi,

I am quite new in the subject and I probably did not properly understood all what a PLS-DA model can do.
But I thought a valided model would be able to classify new individuals with unknown class, which is what I aim for. From what I understood, surely poorly, the predict() function is a cross-validation of the model with individual of known class.
Am I wrong?
I would appreciate some help.

Best regards

hi @LoicHarrault,

Yes you can predict the class of the unknown samples using the function predict() once you have trained the model, see the example here where we pretend we have a new test set: 5 PLS-DA on the SRBCT case study | mixOmics vignette

Kim-Anh

Dear Kim-Anh,

Thank you for your answer. I am trying it.

Best regards

Loïc