Numeric data into categorical

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

Numeric data into categorical

asadbtk
Hello

If we have a dependent variable like yes or no, and independent variables which are continuous, how important is it to transform the independent variables into categorical? Is it something which always improve the accuracy of the machine learning algorithms? I have seen some papers doing these steps. 
Thanks 

_______________________________________________
Wekalist mailing list -- [hidden email]
Send posts to: 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: Numeric data into categorical

Eibe Frank-2
Administrator
No, it does not guarantee improved accuracy. In fact, it may significantly harm accuracy if the discretisation process removes distinguishing features. Having said that, it can help combat overfitting.

Cheers,
Eibe

> On 30/10/2019, at 6:30 AM, javed khan <[hidden email]> wrote:
>
> Hello
>
> If we have a dependent variable like yes or no, and independent variables which are continuous, how important is it to transform the independent variables into categorical? Is it something which always improve the accuracy of the machine learning algorithms? I have seen some papers doing these steps.
> Thanks
> _______________________________________________
> Wekalist mailing list -- [hidden email]
> Send posts to: 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: 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