I presume that you're sampling your data from the "Preprocess" panel which
is not the right way. Instead of that, resmple your instances in the
"Classify" panel. To do so, choose "FilteredClassifier" algorithm.
Left-click on it, select the unsupervised version of the "Resample" filter
for the filter parameter, while for the classifier parameter, select any
method you like.
On 24 September 2016 at 03:56, yasmen [via WEKA] <
I'm working on a new research which i need to
compare my classifier's
performance at different sample sizes while keeping the same number of
classes/categories. However, when using the Weka's Supervised filter
Resample and specifying the sample size to 3% for example, i lose some of
my classes . those who were having a very few number of instances (ex. 1 or
3 records) . i lose them after resampling.
Is there a way to Resample while keeping the classes number the same.
If you reply to this email, your message will be added to the discussion
To unsubscribe from WEKA, click here
View this message in context:
Sent from the WEKA mailing list archive at Nabble.com