prop.txt

*****NEW RESULT*************************
***Prop_B.arff***
best classifier: weka.classifiers.functions.MultilayerPerceptron
arguments: [-L, 0.9016538469375728, -M, 0.36063042122350475, -H, a, -R, -D, -S, 1]
attribute search: null
attribute search arguments: []
attribute evaluation: null
attribute evaluation arguments: []
metric: errorRate
estimated errorRate: 0.1311819662295363
training time on evaluation dataset: 31.297 seconds

You can use the chosen classifier in your own code as follows:

Classifier classifier = AbstractClassifier.forName("weka.classifiers.functions.MultilayerPerceptron", new String[]{"-L", "0.9016538469375728", "-M", "0.36063042122350475", "-H", "a", "-R", "-D", "-S", "1"});
classifier.buildClassifier(instances);


Correctly Classified Instances       20254               86.8004 %
Incorrectly Classified Instances      3080               13.1996 %
Kappa statistic                          0.056 
Mean absolute error                      0.218 
Root mean squared error                  0.3286
Relative absolute error                 94.0518 %
Root relative squared error             96.5349 %
Total Number of Instances            23334     

=== Confusion Matrix ===

     a     b   <-- classified as
 20137    75 |     a = FALSE
  3005   117 |     b = TRUE

=== Detailed Accuracy By Class ===

                 TP Rate  FP Rate  Precision  Recall   F-Measure  MCC      ROC Area  PRC Area  Class
                 0.996    0.963    0.870      0.996    0.929      0.127    0.691     0.929     FALSE
                 0.037    0.004    0.609      0.037    0.071      0.127    0.691     0.283     TRUE
Weighted Avg.    0.868    0.834    0.835      0.868    0.814      0.127    0.691     0.843     

Temporary run directories:
/tmp/autoweka1482182762193476994/


For better performance, try giving Auto-WEKA more time.
Tried 53 configurations; to get good results reliably you may need to allow for trying thousands of configurations.

*********************************************
*****NEW RESULT*************************
***Prop_A.arff***
best classifier: weka.classifiers.trees.J48
arguments: [-S, -M, 12]
attribute search: weka.attributeSelection.GreedyStepwise
attribute search arguments: [-C, -R]
attribute evaluation: weka.attributeSelection.CfsSubsetEval
attribute evaluation arguments: [-L]
metric: errorRate
estimated errorRate: 0.10805500982318271
training time on evaluation dataset: 3.814 seconds

You can use the chosen classifier in your own code as follows:

AttributeSelection as = new AttributeSelection();
ASSearch asSearch = ASSearch.forName("weka.attributeSelection.GreedyStepwise", new String[]{"-C", "-R"});
as.setSearch(asSearch);
ASEvaluation asEval = ASEvaluation.forName("weka.attributeSelection.CfsSubsetEval", new String[]{"-L"});
as.setEvaluator(asEval);
as.SelectAttributes(instances);
instances = as.reduceDimensionality(instances);
Classifier classifier = AbstractClassifier.forName("weka.classifiers.trees.J48", new String[]{"-S", "-M", "12"});
classifier.buildClassifier(instances);


Correctly Classified Instances       41314               89.1945 %
Incorrectly Classified Instances      5005               10.8055 %
Kappa statistic                          0.1971
Mean absolute error                      0.1773
Root mean squared error                  0.2978
Relative absolute error                 85.653  %
Root relative squared error             92.5517 %
Total Number of Instances            46319     

=== Confusion Matrix ===

     a     b   <-- classified as
 40577   310 |     a = FALSE
  4695   737 |     b = TRUE

=== Detailed Accuracy By Class ===

                 TP Rate  FP Rate  Precision  Recall   F-Measure  MCC      ROC Area  PRC Area  Class
                 0.992    0.864    0.896      0.992    0.942      0.277    0.755     0.949     FALSE
                 0.136    0.008    0.704      0.136    0.228      0.277    0.755     0.343     TRUE
Weighted Avg.    0.892    0.764    0.874      0.892    0.858      0.277    0.755     0.878     

Temporary run directories:
/tmp/autoweka4955962859946431316/


For better performance, try giving Auto-WEKA more time.
Tried 45 configurations; to get good results reliably you may need to allow for trying thousands of configurations.

*********************************************
*****NEW RESULT*************************
***Prop_A_O1.arff***
best classifier: weka.classifiers.trees.RandomForest
arguments: [-I, 10, -K, 0, -depth, 0]
attribute search: null
attribute search arguments: []
attribute evaluation: null
attribute evaluation arguments: []
metric: errorRate
estimated errorRate: 0.04382811162472179
training time on evaluation dataset: 9.024 seconds

You can use the chosen classifier in your own code as follows:

Classifier classifier = AbstractClassifier.forName("weka.classifiers.trees.RandomForest", new String[]{"-I", "10", "-K", "0", "-depth", "0"});
classifier.buildClassifier(instances);


Correctly Classified Instances       78193               95.6209 %
Incorrectly Classified Instances      3581                4.3791 %
Kappa statistic                          0.9124
Mean absolute error                      0.0718
Root mean squared error                  0.1833
Relative absolute error                 14.3626 %
Root relative squared error             36.6631 %
Total Number of Instances            81774     

=== Confusion Matrix ===

     a     b   <-- classified as
 37974  2913 |     a = FALSE
   668 40219 |     b = TRUE

=== Detailed Accuracy By Class ===

                 TP Rate  FP Rate  Precision  Recall   F-Measure  MCC      ROC Area  PRC Area  Class
                 0.929    0.016    0.983      0.929    0.955      0.914    0.992     0.993     FALSE
                 0.984    0.071    0.932      0.984    0.957      0.914    0.992     0.991     TRUE
Weighted Avg.    0.956    0.044    0.958      0.956    0.956      0.914    0.992     0.992     

Temporary run directories:
/tmp/autoweka1698543416271728021/


For better performance, try giving Auto-WEKA more time.
Tried 23 configurations; to get good results reliably you may need to allow for trying thousands of configurations.

*********************************************
*****NEW RESULT*************************
***Prop_A_S1.arff***
best classifier: weka.classifiers.trees.RandomForest
arguments: [-I, 10, -K, 0, -depth, 0]
attribute search: null
attribute search arguments: []
attribute evaluation: null
attribute evaluation arguments: []
metric: errorRate
estimated errorRate: 0.03337246557585541
training time on evaluation dataset: 5.133 seconds

You can use the chosen classifier in your own code as follows:

Classifier classifier = AbstractClassifier.forName("weka.classifiers.trees.RandomForest", new String[]{"-I", "10", "-K", "0", "-depth", "0"});
classifier.buildClassifier(instances);


Correctly Classified Instances       78050               95.446  %
Incorrectly Classified Instances      3724                4.554  %
Kappa statistic                          0.9089
Mean absolute error                      0.0858
Root mean squared error                  0.1888
Relative absolute error                 17.1559 %
Root relative squared error             37.765  %
Total Number of Instances            81774     

=== Confusion Matrix ===

     a     b   <-- classified as
 38988  1899 |     a = FALSE
  1825 39062 |     b = TRUE

=== Detailed Accuracy By Class ===

                 TP Rate  FP Rate  Precision  Recall   F-Measure  MCC      ROC Area  PRC Area  Class
                 0.954    0.045    0.955      0.954    0.954      0.909    0.992     0.992     FALSE
                 0.955    0.046    0.954      0.955    0.955      0.909    0.992     0.992     TRUE
Weighted Avg.    0.954    0.046    0.954      0.954    0.954      0.909    0.992     0.992     

Temporary run directories:
/tmp/autoweka6359626463529738897/


For better performance, try giving Auto-WEKA more time.
Tried 23 configurations; to get good results reliably you may need to allow for trying thousands of configurations.

*********************************************
*****NEW RESULT*************************
***Prop_A_U1.arff***
best classifier: weka.classifiers.trees.RandomForest
arguments: [-I, 10, -K, 0, -depth, 0]
attribute search: null
attribute search arguments: []
attribute evaluation: null
attribute evaluation arguments: []
metric: errorRate
estimated errorRate: 0.050994108983799705
training time on evaluation dataset: 0.568 seconds

You can use the chosen classifier in your own code as follows:

Classifier classifier = AbstractClassifier.forName("weka.classifiers.trees.RandomForest", new String[]{"-I", "10", "-K", "0", "-depth", "0"});
classifier.buildClassifier(instances);


Correctly Classified Instances       10281               94.6337 %
Incorrectly Classified Instances       583                5.3663 %
Kappa statistic                          0.8927
Mean absolute error                      0.1406
Root mean squared error                  0.2246
Relative absolute error                 28.1138 %
Root relative squared error             44.9283 %
Total Number of Instances            10864     

=== Confusion Matrix ===

    a    b   <-- classified as
 5121  311 |    a = FALSE
  272 5160 |    b = TRUE

=== Detailed Accuracy By Class ===

                 TP Rate  FP Rate  Precision  Recall   F-Measure  MCC      ROC Area  PRC Area  Class
                 0.943    0.050    0.950      0.943    0.946      0.893    0.988     0.989     FALSE
                 0.950    0.057    0.943      0.950    0.947      0.893    0.988     0.988     TRUE
Weighted Avg.    0.946    0.054    0.946      0.946    0.946      0.893    0.988     0.988     

Temporary run directories:
/tmp/autoweka146212723339239960/


For better performance, try giving Auto-WEKA more time.
Tried 52 configurations; to get good results reliably you may need to allow for trying thousands of configurations.

*********************************************