poi.txt

*****NEW RESULT*************************
***Poi_B.arff***
best classifier: weka.classifiers.meta.RandomSubSpace
arguments: [-I, 36, -P, 0.2757331204049802, -S, 1, -W, weka.classifiers.lazy.IBk, --, -E, -K, 10, -X]
attribute search: null
attribute search arguments: []
attribute evaluation: null
attribute evaluation arguments: []
metric: errorRate
estimated errorRate: 0.08874458874458875
training time on evaluation dataset: 0.213 seconds

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

Classifier classifier = AbstractClassifier.forName("weka.classifiers.meta.RandomSubSpace", new String[]{"-I", "36", "-P", "0.2757331204049802", "-S", "1", "-W", "weka.classifiers.lazy.IBk", "--", "-E", "-K", "10", "-X"});
classifier.buildClassifier(instances);


Correctly Classified Instances         390               84.4156 %
Incorrectly Classified Instances        72               15.5844 %
Kappa statistic                          0.6543
Mean absolute error                      0.295 
Root mean squared error                  0.3481
Relative absolute error                 63.8824 %
Root relative squared error             72.4602 %
Total Number of Instances              462     

=== Confusion Matrix ===

   a   b   <-- classified as
 268  27 |   a = TRUE
  45 122 |   b = FALSE

=== Detailed Accuracy By Class ===

                 TP Rate  FP Rate  Precision  Recall   F-Measure  MCC      ROC Area  PRC Area  Class
                 0.908    0.269    0.856      0.908    0.882      0.657    0.927     0.962     TRUE
                 0.731    0.092    0.819      0.731    0.772      0.657    0.927     0.852     FALSE
Weighted Avg.    0.844    0.205    0.843      0.844    0.842      0.657    0.927     0.922     

Temporary run directories:
/tmp/autoweka2182067311970244071/


For better performance, try giving Auto-WEKA more time.
Tried 258 configurations; to get good results reliably you may need to allow for trying thousands of configurations.

*********************************************
*****NEW RESULT*************************
***Poi_A.arff***
best classifier: weka.classifiers.lazy.LWL
arguments: [-A, weka.core.neighboursearch.LinearNNSearch, -W, weka.classifiers.bayes.BayesNet, --, -Q, weka.classifiers.bayes.net.search.local.K2]
attribute search: weka.attributeSelection.GreedyStepwise
attribute search arguments: [-B, -R]
attribute evaluation: weka.attributeSelection.CfsSubsetEval
attribute evaluation arguments: [-L]
metric: errorRate
estimated errorRate: 0.3078602620087336
training time on evaluation dataset: 0.001 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[]{"-B", "-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.lazy.LWL", new String[]{"-A", "weka.core.neighboursearch.LinearNNSearch", "-W", "weka.classifiers.bayes.BayesNet", "--", "-Q", "weka.classifiers.bayes.net.search.local.K2"});
classifier.buildClassifier(instances);


Correctly Classified Instances         634               69.214  %
Incorrectly Classified Instances       282               30.786  %
Kappa statistic                          0.398 
Mean absolute error                      0.3809
Root mean squared error                  0.4371
Relative absolute error                 76.948  %
Root relative squared error             87.8576 %
Total Number of Instances              916     

=== Confusion Matrix ===

   a   b   <-- classified as
 346  66 |   a = TRUE
 216 288 |   b = FALSE

=== Detailed Accuracy By Class ===

                 TP Rate  FP Rate  Precision  Recall   F-Measure  MCC      ROC Area  PRC Area  Class
                 0.840    0.429    0.616      0.840    0.710      0.420    0.761     0.689     TRUE
                 0.571    0.160    0.814      0.571    0.671      0.420    0.761     0.827     FALSE
Weighted Avg.    0.692    0.281    0.725      0.692    0.689      0.420    0.761     0.765     

Temporary run directories:
/tmp/autoweka7417712631774525933/


For better performance, try giving Auto-WEKA more time.
Tried 299 configurations; to get good results reliably you may need to allow for trying thousands of configurations.

*********************************************
*****NEW RESULT*************************
***Poi_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: 0.14682539682539683
training time on evaluation dataset: 0.209 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         854               84.7222 %
Incorrectly Classified Instances       154               15.2778 %
Kappa statistic                          0.6944
Mean absolute error                      0.2062
Root mean squared error                  0.3119
Relative absolute error                 41.2416 %
Root relative squared error             62.3859 %
Total Number of Instances             1008     

=== Confusion Matrix ===

   a   b   <-- classified as
 453  51 |   a = TRUE
 103 401 |   b = FALSE

=== Detailed Accuracy By Class ===

                 TP Rate  FP Rate  Precision  Recall   F-Measure  MCC      ROC Area  PRC Area  Class
                 0.899    0.204    0.815      0.899    0.855      0.698    0.943     0.944     TRUE
                 0.796    0.101    0.887      0.796    0.839      0.698    0.943     0.948     FALSE
Weighted Avg.    0.847    0.153    0.851      0.847    0.847      0.698    0.943     0.946     

Temporary run directories:
/tmp/autoweka2435631168417345548/


For better performance, try giving Auto-WEKA more time.
Tried 237 configurations; to get good results reliably you may need to allow for trying thousands of configurations.

*********************************************
*****NEW RESULT*************************
***Poi_A_S1.arff***
best classifier: weka.classifiers.rules.DecisionTable
arguments: [-E, acc, -I, -S, weka.attributeSelection.GreedyStepwise, -X, 1]
attribute search: null
attribute search arguments: []
attribute evaluation: null
attribute evaluation arguments: []
metric: errorRate
estimated errorRate: 0.30257936507936506
training time on evaluation dataset: 0.221 seconds

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

Classifier classifier = AbstractClassifier.forName("weka.classifiers.rules.DecisionTable", new String[]{"-E", "acc", "-I", "-S", "weka.attributeSelection.GreedyStepwise", "-X", "1"});
classifier.buildClassifier(instances);


Correctly Classified Instances         706               70.0397 %
Incorrectly Classified Instances       302               29.9603 %
Kappa statistic                          0.4008
Mean absolute error                      0.3972
Root mean squared error                  0.4418
Relative absolute error                 79.4369 %
Root relative squared error             88.3663 %
Total Number of Instances             1008     

=== Confusion Matrix ===

   a   b   <-- classified as
 434  70 |   a = TRUE
 232 272 |   b = FALSE

=== Detailed Accuracy By Class ===

                 TP Rate  FP Rate  Precision  Recall   F-Measure  MCC      ROC Area  PRC Area  Class
                 0.861    0.460    0.652      0.861    0.742      0.423    0.751     0.684     TRUE
                 0.540    0.139    0.795      0.540    0.643      0.423    0.751     0.767     FALSE
Weighted Avg.    0.700    0.300    0.723      0.700    0.692      0.423    0.751     0.725     

Temporary run directories:
/tmp/autoweka2336637252519487324/


For better performance, try giving Auto-WEKA more time.
Tried 387 configurations; to get good results reliably you may need to allow for trying thousands of configurations.

*********************************************
*****NEW RESULT*************************
***Poi_A_U1.arff***
best classifier: weka.classifiers.functions.SMO
arguments: [-C, 1.152106262167039, -N, 0, -M, -K, weka.classifiers.functions.supportVector.Puk -S 5.350239851811477 -O 0.16585023869544807]
attribute search: null
attribute search arguments: []
attribute evaluation: null
attribute evaluation arguments: []
metric: errorRate
estimated errorRate: 0.3300970873786408
training time on evaluation dataset: 0.564 seconds

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

Classifier classifier = AbstractClassifier.forName("weka.classifiers.functions.SMO", new String[]{"-C", "1.152106262167039", "-N", "0", "-M", "-K", "weka.classifiers.functions.supportVector.Puk -S 5.350239851811477 -O 0.16585023869544807"});
classifier.buildClassifier(instances);


Correctly Classified Instances         552               66.9903 %
Incorrectly Classified Instances       272               33.0097 %
Kappa statistic                          0.3398
Mean absolute error                      0.4297
Root mean squared error                  0.4638
Relative absolute error                 85.948  %
Root relative squared error             92.7523 %
Total Number of Instances              824     

=== Confusion Matrix ===

   a   b   <-- classified as
 324  88 |   a = TRUE
 184 228 |   b = FALSE

=== Detailed Accuracy By Class ===

                 TP Rate  FP Rate  Precision  Recall   F-Measure  MCC      ROC Area  PRC Area  Class
                 0.786    0.447    0.638      0.786    0.704      0.349    0.713     0.670     TRUE
                 0.553    0.214    0.722      0.553    0.626      0.349    0.713     0.731     FALSE
Weighted Avg.    0.670    0.330    0.680      0.670    0.665      0.349    0.713     0.701     

Temporary run directories:
/tmp/autoweka6846813057721977537/


For better performance, try giving Auto-WEKA more time.
Tried 327 configurations; to get good results reliably you may need to allow for trying thousands of configurations.

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