Errors using RegressionbyDiscretization [Weka 3.8.5]

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

Errors using RegressionbyDiscretization [Weka 3.8.5]

falete
I'm predicting few steps ahead using a numerical dataset. Multiple algorithms
work well, for instance, SMOreg. However, when I'm using this algorithm
within RegressionbyDiscretization, it fails. Which could be the cause?

I've tried to call many algorithms within RbyD but I got similar errors such
as: "weka cannot handle missing values". I don't know why using the same
dataset fails using the same algorithm.



--
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: Errors using RegressionbyDiscretization [Weka 3.8.5]

Eibe Frank-2
Administrator
RegressionByDiscretization is designed to make classification methods applicable to regression. You can only apply it with a classification method as the base learner. All “Classifier” instances in WEKA that support the “nominal class” capability will work as base learners in RegressionByDiscretization.

Instead of using SMOreg (which can perform regression natively without having to go through RegressionByDiscretization), use SMO (which performs classification only).

Of course, you could go RegressionByDiscretization -> ClassificationViaRegression -> SMOreg (or some other regression method), but that is probably not a good way to proceed in general.

Cheers,
Eibe

> On 12/02/2021, at 5:08 AM, falete <[hidden email]> wrote:
>
> I'm predicting few steps ahead using a numerical dataset. Multiple algorithms
> work well, for instance, SMOreg. However, when I'm using this algorithm
> within RegressionbyDiscretization, it fails. Which could be the cause?
>
> I've tried to call many algorithms within RbyD but I got similar errors such
> as: "weka cannot handle missing values". I don't know why using the same
> dataset fails using the same algorithm.
>
>
>
> --
> 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
_______________________________________________
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: Errors using RegressionbyDiscretization [Weka 3.8.5]

falete
I've tried to convert the numerical attributes to nominal, however, when I go
to forecast tab (Explorer) to apply RegressionbyDiscretization, the
attributes don't appear to be "Target selected". Where should I use
RegreassionbyDiscretization? From 6 attributes I'm only interested on
forecasting 2 of them (my dataset has not classes).





--
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: Errors using RegressionbyDiscretization [Weka 3.8.5]

rparada
I couldn't solve this issue, someone knows how to approach it? thanks



--
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