I registered on mloss.org and am in the process of writing up a
description.
On Sep 29, 2008, at 4:17 PM, Chuck Lam wrote:
I had already planned on going to the conference for other reasons,
so I can
be kind of a "messenger" if none of you can make it.
Of course, ideally one of you can go and actually participate in the
discussion.
On Mon, Sep 29, 2008 at 12:48 PM, Grant Ingersoll
<[EMAIL PROTECTED]>wrote:
Yeah, we need to hurry up, if we're going to do it.
On Sep 29, 2008, at 3:17 PM, Ted Dunning wrote:
Grant,
I could "force" myself to go if necessary as long as it doesn't
conflict
with December 2-3.
Note the October 1 deadline.
On Mon, Sep 29, 2008 at 4:34 AM, Grant Ingersoll
<[EMAIL PROTECTED]
wrote:
I'm interested, but don't know if I can attend, as much as I would
love
to
go to Whistler in December. Let's talk off list, and we can add
others
if
they are interested. We should have a release out w/in the week,
so I
would
think that would be acceptable.
-Grant
On Sep 28, 2008, at 9:47 PM, Chuck Lam wrote:
Hi all,
I've been a lurker on the dev mailing list, hoping to make some
code
contribution at some point but haven't gotten around to it yet :P
In any case, my background is a machine learning researcher, and
I want
the
Mahout developers to be aware of this NIPS Workshop on Machine
Learning
Open
Source Software in December of this year.
http://mloss.org/workshop/nips08/
NIPS is one of the top 3-4 conferences for machine learning
researchers.
It
was in fact at this conference in 2006 where the Stanford
MapReduce
machine
learning paper was presented. (And obviously that paper is an
important
reference point for Mahout.) I think it's extremely important
for Mahout
to
have a presence at the workshop, both to recruit contributors
and to
understand the open source direction of the general machine
learning
community.
The submission deadline is Oct. 1, which obviously is extremely
tight.
However, they're only looking for a 500-word abstract right now,
and
deadline for this type of workshop is usually flexible if we can
convince
the workshop organizers the importance of our work. Another
requirement
is
that the software must be officially released at the time of
submission.
I
don't know the legal definition of that, but I would think that
release
candidate 3 is probably good enough.
I had already planned on attending NIPS, so I'm more than happy
to help
one
of the main Mahout contributors on positioning and presenting
Mahout to
the
NIPS researchers.
--
