Basic question:
When using a trained SVM to classify a test point, am
I right in thinking that one is evaluating a decision
function which involves computing the inner product
between the support vectors from the training phase
and the test point? That is, the SVs are not discarded
after the training stage.
Thanks,
TK
__________________________________________________
Do You Yahoo!?
Everything you'll ever need on one web page
from News and Sport to Email and Music Charts
http://uk.my.yahoo.com
|