| clear query|facets|time |
Search criteria: .
Results from 91 to 100 from
697 (3.078s).
|
|
|
Loading phrases to help you
refine your search...
|
|
Re: Top-N recommendations from SVD - Mahout - [mail # user]
|
|
...I already committed a fix in that direction. I modified our FixedSizePriorityQueue to allow inspection of its head for direct comparison. This obviates the need to instantiate a Comparable a...
|
|
|
Author: Sebastian Schelter,
2013-03-06, 16:07
|
|
|
Re: Top-N recommendations from SVD - Mahout - [mail # user]
|
|
...I tried to rework and optimize the code a little in trunk today, https://issues.apache.org/jira/browse/MAHOUT-1151 You could use this as a basis for further optimization. Best, S...
|
|
|
Author: Sebastian Schelter,
2013-03-06, 12:23
|
|
|
Re: Top-N recommendations from SVD - Mahout - [mail # user]
|
|
...Btw, all important jobs in ALS are map-only, so its the number of map slotes that counts. On 06.03.2013 12:11, Sean Owen wrote:...
|
|
|
Author: Sebastian Schelter,
2013-03-06, 11:16
|
|
|
Re: Top-N recommendations from SVD - Mahout - [mail # user]
|
|
...Looking forward to your debugging results. Could you have a look at the GC behavior? Maybe we should remove the per-tuple object instantiation in line 153: topKItems.offer(new GenericR...
|
|
|
Author: Sebastian Schelter,
2013-03-06, 11:00
|
|
|
Re: Top-N recommendations from SVD - Mahout - [mail # user]
|
|
...Hi Josh, The factorization should be quite a bit faster with the current trunk, as we reworked the QR decomposition used for solving the least squares problems of ALS. I think we...
|
|
|
Author: Sebastian Schelter,
2013-03-06, 10:29
|
|
|
Re: Using ALS job to extract the decomposed matrices - Mahout - [mail # user]
|
|
...The difference is that the job used a reduce-side join to join feature vectors and ratings in 0.5 which is scalable but very slow. We changed this to a broadcast join in later versions...
|
|
|
Author: Sebastian Schelter,
2013-02-27, 10:31
|
|
|
Re: Using ALS job to extract the decomposed matrices - Mahout - [mail # user]
|
|
...Hell Razon, this a strange bug that should not happen. It seems that some of the vectors supplied to the solver are null. Are you sure that there no exceptions previous to this one? &n...
|
|
|
Author: Sebastian Schelter,
2013-02-27, 09:55
|
|
|
Re: 0.8? - Mahout - [mail # dev]
|
|
...Fine for me as well. On 04.02.2013 05:08, Jake Mannix wrote:...
|
|
|
Author: Sebastian Schelter,
2013-02-04, 07:25
|
|
|
Re: 0.8? - Mahout - [mail # dev]
|
|
...I also think that 0.8 should include the new clustering stuff, I recall we wanted even release numbers to contain new features. I plan a hack evening in Berlin with Isabel and Zeno (who port...
|
|
|
Author: Sebastian Schelter,
2013-02-02, 10:33
|
|
|
[MAHOUT-1144] Wrong normalization in SVD++ - Mahout - [issue]
|
|
...Reported by Agnonchik:Seems that I have found a discrepancy of this implementation from the original Yehuda Koren's SVD++ algorithm.line 140:double denominator = Math.sqrt(itemsByUser.size()...
|
|
|
http://issues.apache.org/jira/browse/MAHOUT-1144
Author: Sebastian Schelter,
2013-01-31, 11:15
|
|
|
|
