Is linear discriminant analysis, specifically Linear Programming Discriminant Analysis (LPDA), supervised learning? Can you provide a valid reference that states so if possible.
My study supervisor and I have been disagreeing about it. I'm convinced linear discriminant analysis, whether Fisher LDA or LPDA, is supervised learning. Both techniques use a labelled set of objects to derive a function which can be used to predict class labels for unlabelled objects.
My study supervisor does not agree, stating that nothing is "learned" when using discriminant analysis.