Producer’s accuracy and users’ accuracy

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Producer’s accuracy and users’ accuracy

ivanfrustrado
Hello, first of all, thanks for all your help in the past. And sorry if this is a stupid question. Almost everything is new to me so I have a very basic idea of the concepts and might be the wrong meaning, however, I was showing my supervisor some results of my experiments and he asked for the values of producer’s accuracy and users’ accuracy. In weka, you get a result like this:

Correctly Classified Instances         539               89.5349 %

Incorrectly Classified Instances        63               10.4651 %

Kappa statistic                          0.8604

Mean absolute error                      0.1333

Root mean squared error                  0.2189

Relative absolute error                 35.5459 %

Root relative squared error             50.548  %

Total Number of Instances              602    

 

=== Detailed Accuracy By Class ===

 

                 TP Rate  FP Rate  Precision  Recall   F-Measure  MCC      ROC Area  PRC Area  Class

                 0.834    0.038    0.881      0.834    0.857      0.812    0.975     0.936     1

                 0.941    0.058    0.846      0.941    0.891      0.854    0.987     0.962     2

                 0.847    0.029    0.907      0.847    0.876      0.837    0.977     0.940     3

                 0.960    0.015    0.953      0.960    0.957      0.942    0.997     0.993     4

Weighted Avg.    0.895    0.035    0.897      0.895    0.895      0.861    0.984     0.958    

 

=== Confusion Matrix ===

 

   a   b   c   d   <-- classified as

 126   9  13   3 |   a = 1

   8 143   0   1 |   b = 2

   9  11 127   3 |   c = 3

   0   6   0 143 |   d = 4



so my question is where can I see these values? The porducer's and user's accuracy. Thanks In advance.


Regards. 

Ivan Ruiz.


_______________________________________________
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
|  
Report Content as Inappropriate

Re: Producer’s accuracy and users’ accuracy

Eibe Frank-2
Administrator
Never heard of those terms before in this context. Are they perhaps related to the meaning of the class values?

Cheers,
Eibe

> On 9/06/2017, at 4:03 PM, Ivan Ruiz <[hidden email]> wrote:
>
> Hello, first of all, thanks for all your help in the past. And sorry if this is a stupid question. Almost everything is new to me so I have a very basic idea of the concepts and might be the wrong meaning, however, I was showing my supervisor some results of my experiments and he asked for the values of producer’s accuracy and users’ accuracy. In weka, you get a result like this:
>
> Correctly Classified Instances         539               89.5349 %
>
> Incorrectly Classified Instances        63               10.4651 %
>
> Kappa statistic                          0.8604
>
> Mean absolute error                      0.1333
>
> Root mean squared error                  0.2189
>
> Relative absolute error                 35.5459 %
>
> Root relative squared error             50.548  %
>
> Total Number of Instances              602    
>
>  
> === Detailed Accuracy By Class ===
>
>  
>                  TP Rate  FP Rate  Precision  Recall   F-Measure  MCC      ROC Area  PRC Area  Class
>
>                  0.834    0.038    0.881      0.834    0.857      0.812    0.975     0.936     1
>
>                  0.941    0.058    0.846      0.941    0.891      0.854    0.987     0.962     2
>
>                  0.847    0.029    0.907      0.847    0.876      0.837    0.977     0.940     3
>
>                  0.960    0.015    0.953      0.960    0.957      0.942    0.997     0.993     4
>
> Weighted Avg.    0.895    0.035    0.897      0.895    0.895      0.861    0.984     0.958    
>
>  
> === Confusion Matrix ===
>
>  
>    a   b   c   d   <-- classified as
>
>  126   9  13   3 |   a = 1
>
>    8 143   0   1 |   b = 2
>
>    9  11 127   3 |   c = 3
>
>    0   6   0 143 |   d = 4
>
>
>
>
>
> so my question is where can I see these values? The porducer's and user's accuracy. Thanks In advance.
>
>
>
> Regards.
>
> Ivan Ruiz.
>
> _______________________________________________
> 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
|  
Report Content as Inappropriate

Re: Producer’s accuracy and users’ accuracy

ivanfrustrado
I believe it is for the confusion matrix, like in the first row would be 126/151 * 100  = 83.44% user's accuracy and in the first row would be 126/143 *100 = 88.11% or like in weka the TPRATE and precision, but it does not make any sense to me, like why do you need it. Sorry if I make all this more complex.

2017-06-09 12:11 GMT+08:00 Eibe Frank <[hidden email]>:
Never heard of those terms before in this context. Are they perhaps related to the meaning of the class values?

Cheers,
Eibe

> On 9/06/2017, at 4:03 PM, Ivan Ruiz <[hidden email]> wrote:
>
> Hello, first of all, thanks for all your help in the past. And sorry if this is a stupid question. Almost everything is new to me so I have a very basic idea of the concepts and might be the wrong meaning, however, I was showing my supervisor some results of my experiments and he asked for the values of producer’s accuracy and users’ accuracy. In weka, you get a result like this:
>
> Correctly Classified Instances         539               89.5349 %
>
> Incorrectly Classified Instances        63               10.4651 %
>
> Kappa statistic                          0.8604
>
> Mean absolute error                      0.1333
>
> Root mean squared error                  0.2189
>
> Relative absolute error                 35.5459 %
>
> Root relative squared error             50.548  %
>
> Total Number of Instances              602
>
>
> === Detailed Accuracy By Class ===
>
>
>                  TP Rate  FP Rate  Precision  Recall   F-Measure  MCC      ROC Area  PRC Area  Class
>
>                  0.834    0.038    0.881      0.834    0.857      0.812    0.975     0.936     1
>
>                  0.941    0.058    0.846      0.941    0.891      0.854    0.987     0.962     2
>
>                  0.847    0.029    0.907      0.847    0.876      0.837    0.977     0.940     3
>
>                  0.960    0.015    0.953      0.960    0.957      0.942    0.997     0.993     4
>
> Weighted Avg.    0.895    0.035    0.897      0.895    0.895      0.861    0.984     0.958
>
>
> === Confusion Matrix ===
>
>
>    a   b   c   d   <-- classified as
>
>  126   9  13   3 |   a = 1
>
>    8 143   0   1 |   b = 2
>
>    9  11 127   3 |   c = 3
>
>    0   6   0 143 |   d = 4
>
>
>
>
>
> so my question is where can I see these values? The porducer's and user's accuracy. Thanks In advance.
>
>
>
> Regards.
>
> Ivan Ruiz.
>
> _______________________________________________
> 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


_______________________________________________
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
Loading...