On 05/07/2012 08:31 PM, Darren Govoni wrote:
> Thanks Andy. I'll look at it and perhaps try to complete it. I think
> scikit and ipython are natural fit in this age of cloud computing.
I think the main question is at which level you want to parallelize.
In my experience it pays most to parallelize for model selection,
as you have to do that any way and it is so easy to do.
Certain models, such as forests, are also naturally parallel.
If you want to do a gradient descent in parallel using multiple boxes
(or even CPUs), that's a completely different story and probably
quite hard to do right.


------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general

Reply via email to