mahout-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Razon, Oren" <oren.ra...@intel.com>
Subject "features" parameter in TrainLogistic
Date Tue, 12 Mar 2013 10:42:23 GMT
Hi,
I'm using the "TrainLogistic" class in mahout 0.5 to build a logistic regression.
I'm trying it with the Iris dataset and I do not understand how to work with the "features"
parameter.

It seem that if I give it small int value (~10) I get a coefficient set as a result that contain
the same coefficient value for 2 or more different coefficients.
If I give it a high int value (~>1000) I get a different value per coefficient. What does
this model parameter represent exactly, what value should I give it?

Thanks,
Oren

---------------------------------------------------------------------
Intel Electronics Ltd.

This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message