multiple selection in consequence

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

multiple selection in consequence

Abdrahman0x
Hi...

Is there a way in Weka to make a serial piping filtering (selection). Thus,
can I apply for example an x selection algorithm, then take the results of
that x algorithm and input them into a y algorithm. What I want to do is to
apply filtering of multiple algorithms but the result of each one (output),
will be in input to the second one and so on. Is there a way to implement
that in Weka, and if it is there can anyone explain how to perform it
please.

Thank you,
AR



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

Re: multiple selection in consequence

Peter Reutemann-3
On August 25, 2019 4:02:26 PM GMT+10:00, Abdrahman0x <[hidden email]> wrote:

>Hi...
>
>Is there a way in Weka to make a serial piping filtering (selection).
>Thus,
>can I apply for example an x selection algorithm, then take the results
>of
>that x algorithm and input them into a y algorithm. What I want to do
>is to
>apply filtering of multiple algorithms but the result of each one
>(output),
>will be in input to the second one and so on. Is there a way to
>implement
>that in Weka, and if it is there can anyone explain how to perform it
>please.
>
>Thank you,
>AR
>
>
>
>--
>Sent from: https://weka.8497.n7.nabble.com/
>_______________________________________________
>Wekalist mailing list
>Send posts to: [hidden email]
>To subscribe, unsubscribe, etc., visit
>https://list.waikato.ac.nz/mailman/listinfo/wekalist
>List etiquette:
>http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

Just use the MultiFilter and specify several AttributeSelection filters with your choice of search/evaluation combination.

Cheers, Peter
--
Peter Reutemann
Dept. of Computer Science
University of Waikato, NZ
+64 (7) 858-5174
http://www.cms.waikato.ac.nz/~fracpete/
http://www.data-mining.co.nz/
_______________________________________________
Wekalist mailing list
Send posts to: [hidden email]
To subscribe, unsubscribe, etc., visit https://list.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html
Reply | Threaded
Open this post in threaded view
|

Re: multiple selection in consequence

Abdrahman0x
Thank you Peter for your feedback. Is there a way to know the results
(features) of each selection output.
Also, the MultiFilter will be used in the classify panel, is that correct?

Thank you,
AR



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

Re: multiple selection in consequence

Peter Reutemann-3
On August 26, 2019 3:53:58 PM GMT+10:00, Abdrahman0x <[hidden email]> wrote:

>Thank you Peter for your feedback. Is there a way to know the results
>(features) of each selection output.
>Also, the MultiFilter will be used in the classify panel, is that
>correct?
>
>Thank you,
>AR
>
>
>
>--
>Sent from: https://weka.8497.n7.nabble.com/
>_______________________________________________
>Wekalist mailing list
>Send posts to: [hidden email]
>To subscribe, unsubscribe, etc., visit
>https://list.waikato.ac.nz/mailman/listinfo/wekalist
>List etiquette:
>http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

No, the MultiFilter is a black box. You'd have to write your own filter that would store the dataset coming through to disk, placing it between the attribute selection filters.

Yes, you can use the MultiFilter just like any other filter in conjunction with the FilteredClassifier and your choice of base classifier.

Cheers, Peter
--
Peter Reutemann
Dept. of Computer Science
University of Waikato, NZ
+64 (7) 858-5174
http://www.cms.waikato.ac.nz/~fracpete/
http://www.data-mining.co.nz/
_______________________________________________
Wekalist mailing list
Send posts to: [hidden email]
To subscribe, unsubscribe, etc., visit https://list.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html
Reply | Threaded
Open this post in threaded view
|

Re: multiple selection in consequence

Abdrahman0x
Thank you Peter for the valuable information which made the discussion clear
:)

cheers,
AR



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

Re: multiple selection in consequence

Peter Reutemann-3
In reply to this post by Peter Reutemann-3
> No, the MultiFilter is a black box. You'd have to write your own filter that would store the dataset coming through to disk, placing it between the attribute selection filters.

Just a quick update on this...

This afternoon, I quickly put together a little (unofficial) Weka
package with a "dumping" mechanism for filters, classifiers and
clusterers.

This package offers the following schemes:
- weka.filters.Dumper
- weka.classifiers.meta.Dumper
- weka.clusterers.Dumper

Which in turn allow you to choose a dumping scheme. By default, the
"Null" scheme is used, which does nothing. But you can change this to,
e.g., SaveToDisk, which will save the data to the specified file. That
scheme also has the option to append a timestamp to the filename, to
make it work during cross-validation.

Here is the project page of the package (with download links in the
Releases section):
  https://github.com/fracpete/data-dumper-weka-package

The project page also includes an example layout for a MultiFilter pipeline

HTH

Cheers, Peter
--
My Open Source Blog - http://open.fracpete.org
_______________________________________________
Wekalist mailing list -- [hidden email]
Send posts to: To unsubscribe send an email to [hidden email]
To subscribe, unsubscribe, etc., visit %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html