Hello,
I can just offer "sympathy" - had memory problems, too (12 Gb) for a
large data set. Some of this might be related to (un-necessarily?)
deep-copying the instances structure.
By the way, I've tried different GC strategies (jvm switches) as well
as triggering GC in my code - without success :-( .
As a short term solution you could buy additional memory ;-)
regards,
arne
On Mon, Jun 30, 2008 at 12:04 PM, Sebastian Briesemeister
<
briese@...> wrote:
> Dear all,
>
> some of you might have had already memory problems with weka. Especially
> when you start off with more than 10000 features!
>
> I wonder if this is a problem which can be solved quite easily by triggering
> the garbage collector manually in the weka code??
>
> Otherwise an attribute selection with BestFirst search giving an initial set
> of 20000 features uses up to 9 GB RAM!!!
>
> Does anyone has experience with memory usage of RapidMiner?
>
> Cheers and thanks,
> Sebastian
>
>
>
> _______________________________________________
> Wekalist mailing list
>
Wekalist@...
>
https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist>
>
_______________________________________________
Wekalist mailing list
Wekalist@...
https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
