Re: Problem with J48 Decision Tree

It seems to me as your data doesn't contain any useful information.
I was able to reproduce your behavior with a simple CSV file which contains many lines like this:

A	B	C	Class
1	2	3	YES
1	2	3	NO
1	2	3	YES
1	2	3	NO

The attributes in Columns A, B and C don't contain information useful for classification.
The result of the learning with a J48 tree is the same as with your setup: The whole tree gets pruned (deleted) leaving just one node with the majority class, because it cannot find any useful information in the data.

So, I suggest you take a deep look into your data with some of the available visualization methods of Weka.

Best wishes
Jens



-----Ursprüngliche Nachricht-----
Von: Eibe Frank <[email protected]> 
Gesendet: Donnerstag, 15. Juli 2021 04:07
An: Weka machine learning workbench list. <[email protected]>
Betreff: [Wekalist] Re: Problem with J48 Decision Tree

Your tree doesn't use *any* predictor attributes: this tree consists of a single leaf node that simply predicts the majority class value.

Cheers,
Eibe

> On 14/07/2021, at 1:51 AM, [email protected] wrote:
> 
> I keep running the data through and it makes a binary decision, and only appears to use one attribute, which it doesn't specify. What am I missing in configuring the tree? Output is:
> 
> J48 pruned tree
> ------------------
> : High (113.0/57.0)
> 
> Number of Leaves  : 	1
> 
> Size of the tree : 	1
> 
> 
> Time taken to build model: 0 seconds
> 
> === Stratified cross-validation ===
> === Summary ===
> 
> Correctly Classified Instances          56               49.5575 %
> Incorrectly Classified Instances        57               50.4425 %
> Kappa statistic                          0     
> Mean absolute error                      0.4178
> Root mean squared error                  0.4572
> Relative absolute error                 99.8172 %
> Root relative squared error             99.9995 %
> Total Number of Instances              113     
> 
> _______________________________________________
> Wekalist mailing list -- [email protected] Send posts to 
> [email protected] To unsubscribe send an email to 
> [email protected] 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 -- [email protected] Send posts to [email protected] To unsubscribe send an email to [email protected] 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 -- [email protected]
Send posts to [email protected]
To unsubscribe send an email to [email protected]
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