Hi
I also have a similar question regarding result interpretation based on how
we provide data to recommender.
Typcally , we provide rating data say in scale from 1-5 and get the result
in the same scale range.(and need to be consistent as Sean points out)
If we assume the provided data with
error when the
value is large.
Ted I think this is why you say the SVD is bad for count-like data?
On Feb 6, 2013 8:28 AM, 万代豊 20525entrad...@gmail.com wrote:
Hi
I also have a similar question regarding result interpretation based on
how
we provide data to recommender.
Typcally , we
list somewhere.
Regards,,,
Yutaka
iPhoneから送信
On 2013/01/31, at 23:33, Jake Mannix jake.man...@gmail.com wrote:
Hi Yutaka,
On Thu, Jan 31, 2013 at 3:03 AM, 万代豊 20525entrad...@gmail.com wrote:
Hi
Here is a question around how to evaluate the result of Mahout 0.7
This is the agenda that I'm interested in too.
I believe Item-Based Recomemndation in Mahout (Not only about Mahout
though) should spend sometime
doing multiplication of cooccurrence matrix and user preference vector.
If we could pass this multiplication task off loaded to GGPU, then that
will be
-kmeans-clusterdump01
-b 30-n 10
I believe the vectordump utility and the clusterdump derive from different
roots in terms of it's heap requirement.
Still waiting for some advise from you people.
Regards,,,
Y.Mandai
2013/2/19 万代豊 20525entrad...@gmail.com
Well , the --sortVectors for the vectordump
Hi
Is this feature already unsupported in Mahout-0.7?
Some topics tells that this has moved to WeightedVectorWritables, however
still not sure how I can
pull out labels from clusters other than top terms from ClusterDumper
K-Means clustering from vectors created from Lucene index (vi Mahout
source.
Thank you.
Regards,,,
Y.Mandai
iPhoneから送信
On 2013/02/23, at 3:14, Jake Mannix jake.man...@gmail.com wrote:
On Fri, Feb 22, 2013 at 2:26 AM, 万代豊 20525entrad...@gmail.com wrote:
Thanks Jake for your attention on this.
I believe I have the trunk code from the official
Hi all
I guest I must've seen somewhere on very similar topics on classname change
in Mahout-0.8-SNAPSHOT for some of the Lucene analyzer and here is another
one that I need to be solved.
Mahout gave me an error for seq2sparse with Lucene analyzer option as
follows,
which of cource had been
Not sure unless I intetionally reproduce this situation, however, Mahout
recommendation seems to be senstive to the carriage code placed at the end
of your final input data record.
For instance,
If your final data record ends like
5 105
5 106CRLF
End of File
and have no succeeding recods, I
be
org.apache.lucene.analysis.core.WhitespaceAnalyzer ( u were missing the
'core')
Mahout trunk's presently at Lucene 4.2.1. Lucene's has gone through a
major refactor in 4.x.
Check Lucene 4.2.1 docs for the correct package name.
From: 万代豊 20525entrad...@gmail.com
-jobtracker-info-could-only-be-replicated-to-0-nodes-instead-of-1
From: 万代豊 20525entrad...@gmail.com
To: user@mahout.apache.org user@mahout.apache.org
Sent: Saturday, May 11, 2013 1:14 PM
Subject: Re: Class Not Found from 0.8-SNAPSHOT
Hi
After my several successful jobs experiences on other Mahout Kmeans
calculation in the past , I'm facing a sudden heap error as below in Mahout
seq2sparse process.(Mahout-0.70 on Hadoop-0.20.203 Pseudo-distributed)
[hadoop@localhost TEST]$ $MAHOUT_HOME/bin/mahout seq2sparse --namedVector
-i
Hi Shweta
I guess I can handle this.
I always specify namedVector option when generation term vector(seq2sparse)
as;
$MAHOUT_HOME/bin/mahout seq2sparse --namedVector -i MyJob/MyJob-seqfile/ -o
MyJob/MyJob-namedVector -ow -a
org.apache.lucene.analysis.WhitespaceAnalyzer -chunk 200 -wt tfidf -s 5
trying to increase the heap size for child JVMs per this doc,
depending on what version you're running:
http://hadoop.apache.org/docs/r2.5.1/hadoop-project-dist/hadoop-common/ClusterSetup.html
On Tue, Dec 16, 2014 at 11:33 PM, 万代豊 20525entrad...@gmail.com wrote:
Hi
After my several
Hi Jakub
To label the training data for Bayesian classification in Mahout, all you
do is just simply place your text training file into folders with the
desired label as folder names.
For example, in case of 20-news group, you can place your text into
following folders as,
[hadoop@localhost
Hi
Looks like this is typical everywhere, however I have'nt figured out how to
resolve in my case.
There is nothing I have done explicitly regarding SLF4J.
Both Hadoop and Mahout environment are built by just simply downloading jar
files. Not built locally.
Both Hadoop and Mahout have been
16 matches
Mail list logo
