Performance of linear discriminant analysis when the normality assumption is violated

  • $\begingroup$ This is a very broad question. The performance depends on your data but stats.stackexchange.com/questions/110908/… and stats.stackexchange.com/questions/113133/… address this $\endgroup$ – Blue Eyed Daisy Feb 29 at 12:43
  • $\begingroup$ LDA with 2+ classes is a two-stage procedure: (1) extraction of discriminants and (2) classifying objects. (1) assumes symmetric unimodal distribution, such as normal but not necessarily normal. (2) assumes normal because (typically) we rely on normal pdf in this classification (the gaussian classifier). $\endgroup$ – ttnphns Feb 29 at 17:38

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.