[prev in list] [next in list] [prev in thread] [next in thread] 

List:       wekalist
Subject:    [Wekalist] Apriori
From:       <Sebastien.Vouriot () loria ! fr>
Date:       2003-06-30 13:30:50
[Download RAW message or body]

Hi there, 
 
I found some troubles in running the weka system for association rules 
extraction (i.e. the Apriori program). 
 
Did anyone hear or read some mandatory thresholds on the input data. 
 
I ran Apriori with these different test-sets: 
 
-6 instances and 30 attributes = rules extracted succesfully 
-6 instances and 48 attributes = outofmemory  
-9 instances and 80 attributes = rules extracted succesfully 
-9 instances and 82 attributes = outofmemory 
 
on different machines, the strongest one is double-processor 2GigaBytes 
of RAM. 
 
 
I'm wondering what's wrong with my data? Is the ratio (instances x 
attributes) the problem? I suspect that the more instancesXattributes 
matrix rectangular is, the more the process tends to crash. 
 
Thanks for your help, 
Greetings, 
 
Sebastian 


_______________________________________________
Wekalist mailing list
Wekalist@list.scms.waikato.ac.nz
https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
[prev in list] [next in list] [prev in thread] [next in thread]