Simple logistic regression: How do class 0 and class 1 map to my own outcome variables?

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

Simple logistic regression: How do class 0 and class 1 map to my own outcome variables?

surangak

Hi all,

Im using Weka to build decision models using Simple Logistic Regression. My data has two outcome classes, N and P.

But my model outputs scores for each outcomes as class 0 and class 1. 

It also outputs the performance metrics (sensitivity/specificity etc.) and a confusion matrix. But these are defined as N and P (which are what I really have as my outcome variables)


Class 0 :
-0.12 + 
[id_col] * 0    +
[feature_A] * 0.16 +
[feature_C] * 0.04 +
[feature_B] * -0.19 +

Class 1 :
0.12 + 
[id_col] * -0 +
[feature_D] * -0.16 +
[feature_B] * -0.04 +

My question is, what does classes 0 and 1 map to? which maps to N, and which to p?


Best regards,
Suranga


--
Best Regards,
Suranga

_______________________________________________
Wekalist mailing list
Send posts to: [hidden email]
List info and subscription status: 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: Simple logistic regression: How do class 0 and class 1 map to my own outcome variables?

Eibe Frank-2
Administrator
Good point. We should improve the output.

The indexing is consistent with the order of the class values in the class attributes, as shown, for example, in the Preprocess panel of the Explorer.

Cheers,
Eibe

> On 29 Mar 2017, at 05:53, Suranga Kasthurirathne <[hidden email]> wrote:
>
>
> Hi all,
>
> Im using Weka to build decision models using Simple Logistic Regression. My data has two outcome classes, N and P.
>
> But my model outputs scores for each outcomes as class 0 and class 1.
>
> It also outputs the performance metrics (sensitivity/specificity etc.) and a confusion matrix. But these are defined as N and P (which are what I really have as my outcome variables)
>
>
> Class 0 :
> -0.12 +
> [id_col] * 0    +
> [feature_A] * 0.16 +
> [feature_C] * 0.04 +
> [feature_B] * -0.19 +
>
> Class 1 :
> 0.12 +
> [id_col] * -0 +
> [feature_D] * -0.16 +
> [feature_B] * -0.04 +
>
> My question is, what does classes 0 and 1 map to? which maps to N, and which to p?
>
>
> Best regards,
> Suranga
>
>
> --
> Best Regards,
> Suranga
> _______________________________________________
> Wekalist mailing list
> Send posts to: [hidden email]
> List info and subscription status: https://list.waikato.ac.nz/mailman/listinfo/wekalist
> List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

_______________________________________________
Wekalist mailing list
Send posts to: [hidden email]
List info and subscription status: 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: Simple logistic regression: How do class 0 and class 1 map to my own outcome variables?

Eibe Frank-2
Administrator
In reply to this post by surangak
This is fixed now in the main trunk version in the code repository.

Cheers,
Eibe

> On 29 Mar 2017, at 05:53, Suranga Kasthurirathne <[hidden email]> wrote:
>
>
> Hi all,
>
> Im using Weka to build decision models using Simple Logistic Regression. My data has two outcome classes, N and P.
>
> But my model outputs scores for each outcomes as class 0 and class 1.
>
> It also outputs the performance metrics (sensitivity/specificity etc.) and a confusion matrix. But these are defined as N and P (which are what I really have as my outcome variables)
>
>
> Class 0 :
> -0.12 +
> [id_col] * 0    +
> [feature_A] * 0.16 +
> [feature_C] * 0.04 +
> [feature_B] * -0.19 +
>
> Class 1 :
> 0.12 +
> [id_col] * -0 +
> [feature_D] * -0.16 +
> [feature_B] * -0.04 +
>
> My question is, what does classes 0 and 1 map to? which maps to N, and which to p?
>
>
> Best regards,
> Suranga
>
>
> --
> Best Regards,
> Suranga
> _______________________________________________
> Wekalist mailing list
> Send posts to: [hidden email]
> List info and subscription status: https://list.waikato.ac.nz/mailman/listinfo/wekalist
> List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

_______________________________________________
Wekalist mailing list
Send posts to: [hidden email]
List info and subscription status: 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: Simple logistic regression: How do class 0 and class 1 map to my own outcome variables?

surangak

Amazing, thank you! Never say I don't contribute anything to Weka ;)

On Sat, May 6, 2017 at 4:20 AM, Eibe Frank-2 [via WEKA] <[hidden email]> wrote:
This is fixed now in the main trunk version in the code repository.

Cheers,
Eibe

> On 29 Mar 2017, at 05:53, Suranga Kasthurirathne <[hidden email]> wrote:
>
>
> Hi all,

>
> Im using Weka to build decision models using Simple Logistic Regression. My data has two outcome classes, N and P.
>
> But my model outputs scores for each outcomes as class 0 and class 1.
>
> It also outputs the performance metrics (sensitivity/specificity etc.) and a confusion matrix. But these are defined as N and P (which are what I really have as my outcome variables)
>
>
> Class 0 :
> -0.12 +
> [id_col] * 0    +
> [feature_A] * 0.16 +
> [feature_C] * 0.04 +
> [feature_B] * -0.19 +
>
> Class 1 :
> 0.12 +
> [id_col] * -0 +
> [feature_D] * -0.16 +
> [feature_B] * -0.04 +
>
> My question is, what does classes 0 and 1 map to? which maps to N, and which to p?
>
>
> Best regards,
> Suranga
>
>
> --
> Best Regards,
> Suranga
> _______________________________________________
> Wekalist mailing list
> Send posts to: [hidden email]
> List info and subscription status: https://list.waikato.ac.nz/mailman/listinfo/wekalist
> List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

_______________________________________________
Wekalist mailing list
Send posts to: [hidden email]
List info and subscription status: https://list.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html



To unsubscribe from Simple logistic regression: How do class 0 and class 1 map to my own outcome variables?, click here.
NAML



--
Best Regards,
Suranga