Re: Help needed to interpret weka output

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Re: Help needed to interpret weka output

Eibe Frank-3
Those numbers are counts. For example, 103 is the number of instances for which class=Democrat and handicapped-infants=n.

The numbers in brackets give the proportion of each class of instances in the data.

You may have to take the Laplace correction into account when considering theses numbers.

Note that the model shown by default is the model buit from the full dataset (as loaded into the Preprocess panel). It doesn’t matter whether you use cross-validation or percentage split evaluation. The model won’t change. The latest version of WEKA has an option to output all models built during evaluation, under “More options...”.

Cheers,
Eibe

From: Harmeet Singh <[hidden email]>
To: [hidden email]
Cc: 
Bcc: 
Date: Fri, 23 Feb 2018 19:49:21 -0800
Subject: Help needed to interpret weka output
Hi,

I am using the vote.arff dataset to train NaiveBayes classifier in weka. I am using the default values except cross-validation folds (5 instead of default). I get the output in the following format:

Naive Bayes Classifier
TABLE-1
                                              Class
Attribute                                  democrat republican
                                             (0.61)     (0.39)
===============================================================
handicapped-infants
  n                                            103.0      135.0
  y                                            157.0       32.0
  [total]                                      260.0      167.0

water-project-cost-sharing
  n                                            120.0       74.0
  y                                            121.0       76.0
  [total]                                      241.0      150.0

adoption-of-the-budget-resolution
  n                                             30.0      143.0
  y                                            232.0       23.0
  [total]                                      262.0      166.0

physician-fee-freeze
  n                                            246.0        3.0
  y                                             15.0      164.0
  [total]                                      261.0      167.0

el-salvador-aid
  n                                            201.0        9.0
  y                                             56.0      158.0
  [total]                                      257.0      167.0

.
.
.
.
.


Time taken to build model: 0 seconds

=== Stratified cross-validation ===
=== Summary ===

Correctly Classified Instances         392               90.1149 %
Incorrectly Classified Instances        43                9.8851 %
Kappa statistic                          0.7953
Mean absolute error                      0.1006
Root mean squared error                  0.2999
Relative absolute error                 21.2108 %
Root relative squared error             61.5991 %
Total Number of Instances              435     

=== Confusion Matrix ===

   a   b   <-- classified as
 237  30 |   a = democrat
  13 155 |   b = republican


My question is:
What does each entry of table-1 represent? Can someone please explain the output format?


Regards,
Harmeet Singh

_______________________________________________
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