Homogeneity of Covariance Matrix and LDA

Asked Jan 23 '17 at 20:15

Active Jan 23 '17 at 20:15

Viewed 562 times

I have been using Iris data for a classification problem in SPSS.

The Box's M-test is used to check the assumption whether all co-variance matrices are equal or not. Now since the p.value for Iris data is equal to 0.000<0.05 criteria which is set by default in SPSS is it possible to still evaluate the LDA problem using the IRIS data?

As far as I know if the assumption for homogeneity between co variance is not satisfied then we have to use an alternative way to analyze our data? Am I correct or we still can analyze our data using LDA?

asked Jan 23 '17 at 20:15

Annalise Azzopardi

Start by reading http://stats.stackexchange.com/a/71571/3277, http://stats.stackexchange.com/a/190821/3277. – ttnphns Jan 24 '17 at 08:15
@ttphns - Yes thank you i will check them out for better understanding – Annalise Azzopardi Jan 29 '17 at 11:34

Homogeneity of Covariance Matrix and LDA

0 Answers0