Hi Isabel,
Can you quickly take some time to validate
my EM understanding in clustering context ?
I have written my idea as comments on MAHOUT-4 before
turning out the new code.
If you (or anyone else) can quickly validate it and
possibly refine/rectify any gaps/mistakes in it, then
I can turn out the non-parallel version along with some
sample test cases quickly.
Thanks
-Ankur
-
