clear query| facets| time Search criteria: author:"Ted Dunning".   Results from 31 to 40 from 5214 (0.0s).
Loading phrases to help you
refine your search...
[expand - 3 more] - Large-scale Language Models - Mahout - [mail # user]
...Yeah... rumor is that version 20 had a huge performance improvement over 19.  Lots of applications became much more feasible at that point.  On Wed, Feb 10, 2010 at 3:52 AM, Isabel...
   Author: Ted Dunning , 2010-02-10, 18:31
iletken paper - Mahout - [mail # user]
...This is a bit out of date and the white paper you review is clearly written from a biased point of view.  Some salient points,  -  Mahout includes a good bit more than just Ta...
   Author: Ted Dunning , 2010-02-10, 23:16
Machine Learning Question - Mahout - [mail # user]
...I think I understand your question.  To make sure, here it is in my terms:  - you have documents with tag tokens in the fid field  - you have a bunch of rules for defining whi...
   Author: Ted Dunning , 2010-02-17, 20:22
[expand - 1 more] - Naive Bayes on Well Structured Data - Mahout - [mail # user]
...This sounds great.  I would suggest you test the naive Bayes, complementary Naive Bayes, SVM and SGD implementations.  Given that naive Bayes has worked well on a sample, you will ...
   Author: Ted Dunning , 2010-02-17, 22:10
[expand - 2 more] - Need a little help with SVD / Dimensional Reduction - Mahout - [mail # user]
...Yes.  These are term vectors, not document vectors.  There is an additional step that can be run to produce document vectors.  On Sun, Jun 5, 2011 at 1:16 PM, Stefan Wienert &...
   Author: Ted Dunning , 2011-06-06, 10:08
[expand - 1 more] - Logistic Regression + Time Series - Mahout - [mail # user]
...What Hector said.  You will need to extract features from your time history.  The question also comes up about how large is  your data set.  If it is less than 100,000 tr...
   Author: Ted Dunning , 2011-06-06, 10:12
How to get the predicted target lable using CrossFolderLearner? - Mahout - [mail # user]
...You have to remember that mapping.  You will have created it when you encoded the target variable.  This is occasionally a nasty problem.  I have considered adding the ability...
   Author: Ted Dunning , 2011-06-06, 10:17
[expand - 2 more] - Computing SVD Of "Large Sparse Data" - Mahout - [mail # user]
...I would push for SSVD as well if you want a real SVD.  Also, I don't think that you lose information about which vectors are which (or as Jake put it "what they mean").  The stocha...
   Author: Ted Dunning , 2011-06-06, 10:32
[expand - 1 more] - What's the best practice to specify df for TPrior and alphaByLambda for ElasticBandPrior - Mahout - [mail # user]
...This paper has good guidance.  I think that their recommendation of df=1 and scale=2.5 is very reasonable.  Their suggestion of mean subtraction would be disastrous, but much of th...
   Author: Ted Dunning , 2011-06-06, 11:39
[expand - 1 more] - Turning Preference Files Into Vectors - Mahout - [mail # user]
...Fantastic.  That will be great.  On Sat, Feb 6, 2010 at 12:05 PM, Matthew Bryan  wrote:  > hopefully I'll soon > get to the point that I'm comfortable contributing ...
   Author: Ted Dunning , 2010-02-07, 00:06