On 7/03/16, 9:19 PM, "khalil ghafoori" <wekalist-bounces(a)list.waikato.ac.nz
on behalf of ghafoori_khalil(a)yahoo.com> wrote:
I'am master student in SRTTU.edu
and study on multilabel classification.
I use Weka for get association rule but in bibtex.arff dataset :
i use -Xmx4G in command line on server machime with 16GB ram.
out message :
OutofMemoryError : GC overhead limit exceede.
pls help me.
You could try increasing the amount of heap space given to the virtual machine. Another
thing to experiment with is the settings for support and confidence. Start with fairly
high values for both in order to limit the number of large item sets and rules considered
by the algorithm. If your data is binary valued, then you could try FPGrowth instead of
APriori. FPGrowth has an additional option to limit the maximum number of items included
in large item sets and rules.