SVM - RBF (Training Split)
80 20 0.767857142857  - 29, 195
70 30 0.797619047619  - 25, 171
60 40 0.830357142857  - 23, 145
50 50 0.814285714286  - 16, 124


------------------------------------

80 - 20
4 depth Decision Tree 0.785714285714
K Neighbors Classifier 0.785714285714
SVM - RBF 0.767857142857
Ensemble 0.785714285714


60 - 40
4 depth Decision Tree 0.821428571429
K Neighbors Classifier 0.848214285714
SVM - RBF 0.830357142857
Ensemble 0.794642857143

50 - 50
4 depth Decision Tree 0.785714285714
K Neighbors Classifier 0.814285714286
SVM - RBF 0.814285714286
Ensemble 0.814285714286

40 - 60
4 depth Decision Tree 0.833333333333
K Neighbors Classifier 0.821428571429
SVM - RBF 0.821428571429
Ensemble 0.821428571429

30 - 70
4 depth Decision Tree 0.780612244898
K Neighbors Classifier 0.841836734694
SVM - RBF 0.841836734694
Ensemble 0.80612244898

20 - 80
4 depth Decision Tree 0.75
K Neighbors Classifier 0.852678571429
SVM - RBF 0.857142857143
Ensemble 0.736607142857


----------------------------------------
LSTM Baselines

Answers
Test score: 1.35040706396
Test categorical accuracy: 0.671875
Test fmeasure: 0.66666662693
Test accuracy: 0.671875


Questions
Test score: 0.00850728852674
Test categorical accuracy: 1.0
Test fmeasure: 1.0
Test accuracy: 1.0


-----------------------------------------
Joint Probability MLP Architecture



-----------------------------------------


Conditional Architecture
------------------------
Conditioned on Question
Answer Accuracy:

('Test score:', 0.58240962028503418)
('Test categorical accuracy:', 0.75)
('Test fmeasure:', 0.74999994039535522)
('Test accuracy:', 0.75)