How to get different results in the classify tab

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

How to get different results in the classify tab

jakiehals
I've started using weka and I have a file that I use in the classify tab with
different algorithms. For each algorithm I want to improve the algorithm so
I can find the maximum accuracy of the result. So I tried changing the batch
size. With the default batch size: 100 the results are these

<https://weka.8497.n7.nabble.com/file/t7285/Capture.png>

when I change the batch size to any other number I get the same results. I
am using the MultiLayerPerceptron algorithm. Is it normal that the numbers
are the same? If not is there anything else I can change to get different
results? Thank you!




--
Sent from: https://weka.8497.n7.nabble.com/
_______________________________________________
Wekalist mailing list -- [hidden email]
Send posts to [hidden email]
To unsubscribe send an email to [hidden email]
To subscribe, unsubscribe, etc., visit https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html
Reply | Threaded
Open this post in threaded view
|

Re: How to get different results in the classify tab

Peter Reutemann
> I've started using weka and I have a file that I use in the classify tab with
> different algorithms. For each algorithm I want to improve the algorithm so
> I can find the maximum accuracy of the result. So I tried changing the batch
> size. With the default batch size: 100 the results are these
>
> <https://weka.8497.n7.nabble.com/file/t7285/Capture.png>
>
> when I change the batch size to any other number I get the same results. I
> am using the MultiLayerPerceptron algorithm. Is it normal that the numbers
> are the same? If not is there anything else I can change to get different
> results? Thank you!

Changing the batch size won't change the results at all. It is merely
a way of bundling several rows of data to generate predictions for
them in one go (some classifiers are more efficient that way).

For parameter optimization use either GridSearch or MultiSearch to
explore an algorithm's parameter space.

Cheers, Peter
--
Peter Reutemann
Dept. of Computer Science
University of Waikato, NZ
+64 (7) 577-5304
http://www.cms.waikato.ac.nz/~fracpete/
http://www.data-mining.co.nz/
_______________________________________________
Wekalist mailing list -- [hidden email]
Send posts to [hidden email]
To unsubscribe send an email to [hidden email]
To subscribe, unsubscribe, etc., visit https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html