Everyone can discuss about the cut-off points in linear discriminant analysis.
I know how to calculate it when there are 2 groups. But, more than 2 groups such as 3-5 groups, is it impossible?

I'm in-charge the application scorecard project and i want to classy customer to 5 groups from initial data set using discriminant analysis and built a discrimination function for new customers and group them by cut-off points.

Thanks

Could you possibly elaborate on cut-off points a little ? It's kind of support vectors, isn't it ?

the cutting score is kind of support vector machines.
Can you take a brief look at the document below about discriminant function analysis and example?
http://www.fileden.com/files/2007/8/3/1320089/discriminant.ppt