| clear query|facets|time |
Search criteria: .
Results from 111 to 120 from
2731 (0.295s).
|
|
|
Loading phrases to help you
refine your search...
|
|
Re: Discossuon Of ML environment/MR, Mahout - Mahout - [mail # dev]
|
|
...I'm trying to separate what a particular implementation happens to do on framework X from what is possible, and with what difficulty. I'm not convinced that this example, maybe others, shows...
|
|
|
Author: Sean Owen,
2013-03-12, 22:21
|
|
|
Re: mahout collections updates - Mahout - [mail # dev]
|
|
...It's almost certainly the overhead of the iterator creation and iterator methods. DenseVector.dot() is not specialized and the simple dot product method here could as well be placed there. T...
|
|
|
Author: Sean Owen,
2013-03-12, 22:00
|
|
|
Re: Problem in Deploying Mahout Recommender As a Web Service - Mahout - [mail # user]
|
|
...Yes, that was removed a long time ago. You can still find the servlets and such in the integration module, and can put together your own web app that way. It's just that the packaging itself...
|
|
|
Author: Sean Owen,
2013-03-12, 10:20
|
|
|
Re: Out-of-core random forest implementation - Mahout - [mail # dev]
|
|
...On Fri, Mar 8, 2013 at 11:54 PM, Ted Dunning wrote: The pattern here seems to be 'checkpointing' as the culprit -- that's well said. In an iterative M/R process you are necessar...
|
|
|
Author: Sean Owen,
2013-03-09, 09:16
|
|
|
Re: Out-of-core random forest implementation - Mahout - [mail # dev]
|
|
...Oh, certainly. I was thinking in the realm of distributed systems only. Surely serialization across a network is a necessary step in anything like that. Serializing to local disk first, or a...
|
|
|
Author: Sean Owen,
2013-03-08, 23:03
|
|
|
Re: Out-of-core random forest implementation - Mahout - [mail # dev]
|
|
...You can get a Hadoop job to finish in wrote: ...
|
|
|
Author: Sean Owen,
2013-03-08, 14:41
|
|
|
Re: Out-of-core random forest implementation - Mahout - [mail # dev]
|
|
...Weighing in late here -- it's an interesting question whether M/R is a good fit for iterative processes. Today you can already optimize away most of the startup overhead for a job, by settin...
|
|
|
Author: Sean Owen,
2013-03-08, 14:03
|
|
|
Re: Top-N recommendations from SVD - Mahout - [mail # user]
|
|
...That too, even better. Isn't that already done? Could be in one place but not another. IIRC there were also cases where it was a lot easier to pass around an object internally and mutability...
|
|
|
Author: Sean Owen,
2013-03-06, 16:05
|
|
|
Re: Top-N recommendations from SVD - Mahout - [mail # user]
|
|
...OK and he mentioned that 10 mappers were running, when it ought to be able to use several per machine. The # of mappers is a function of the input size really, so probably needs to turn down...
|
|
|
Author: Sean Owen,
2013-03-06, 11:49
|
|
|
Re: Top-N recommendations from SVD - Mahout - [mail # user]
|
|
...OK, that's reasonable on 35 machines. (You can turn up to 70 reducers, probably, as most machines can handle 2 reducers at once). I think the recommendation step loads one whole matrix into ...
|
|
|
Author: Sean Owen,
2013-03-06, 11:11
|
|
|
|
