How to determine the labeling class in the metrics of sensitivity/specificity/false-positive

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

How to determine the labeling class in the metrics of sensitivity/specificity/false-positive

sjtu
Hi WEKA experts and users,

I have a quick question on how should I determine what order of labeling class in my CSV file is used to calculate the false-positive/sensitivity/specificity in the 2-class classification.

I have a CSV file and I want to use a classifier to perform the 2-class classification.
The 100 attributes are arranged sequentially in the first row and the labeling class is the end of the first row.
I arrange the first class ("A") in the 2-100 rows and the second class ("B") in the 101-200 rows
I can get the metrics of classification performance in WEKA.

In the 2-class classification, we use positive/negative (true/false) class to  determine the true/false table and then calculate the metrics of false-positive rate/sensitivity/specificity   etc.

My question is does my first labeling class "A" correspond the the positive (true) label in the performance metrics of WEKA?
Or my second labeling class "B" correspond to the positive (true) label in the performance metrics of WEKA?

Warm regards,
Shu-Ju





_______________________________________________
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: How to determine the labeling class in the metrics of sensitivity/specificity/false-positive

Eibe Frank
In the Explorer, etc., the output contains a separate FPR, etc., for each label. Are you using the Experimenter? There is an option in the ClassifierSplitEvaluator (you will need to switch to the advanced mode of the Experimenter) to change the index of the label that is used for "IR" (information retrieval) statistics.

Cheers,
Eibe

On Tue, Feb 11, 2020 at 7:18 PM Shu-Ju Tu <[hidden email]> wrote:
Hi WEKA experts and users,

I have a quick question on how should I determine what order of labeling class in my CSV file is used to calculate the false-positive/sensitivity/specificity in the 2-class classification.

I have a CSV file and I want to use a classifier to perform the 2-class classification.
The 100 attributes are arranged sequentially in the first row and the labeling class is the end of the first row.
I arrange the first class ("A") in the 2-100 rows and the second class ("B") in the 101-200 rows
I can get the metrics of classification performance in WEKA.

In the 2-class classification, we use positive/negative (true/false) class to  determine the true/false table and then calculate the metrics of false-positive rate/sensitivity/specificity   etc.

My question is does my first labeling class "A" correspond the the positive (true) label in the performance metrics of WEKA?
Or my second labeling class "B" correspond to the positive (true) label in the performance metrics of WEKA?

Warm regards,
Shu-Ju




_______________________________________________
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