How can I do a classifer in LDA (manual)

Asked Sep 05 '12 at 16:04

Active Sep 11 '12 at 11:56

Viewed 314 times

I'm trying to make the classification rule of LDA in R , this is using the Euclidean's distance, g(x)= t(w)x - wo, w is my eigenvector, x my test data, wo the mean of the two classes.

My question is, how can I pass the model (project data) to model prediction (True-False)? thank you

edited Sep 06 '12 at 09:55

Sicco

6,167
5
45
61

asked Sep 05 '12 at 16:04

Jessica Medina

Have you looked at the MASS package? Specifically `lda` and `predict.lda`. – Brandon Bertelsen Sep 05 '12 at 17:43
yes, I know this package, but I need do it by hand – Jessica Medina Sep 05 '12 at 19:09
The package source would be instructive in that case. – Brandon Bertelsen Sep 06 '12 at 05:47

How can I do a classifer in LDA (manual)

0 Answers0