Sample Output.txt

C:\Users\James\Anaconda3\envs\keras-gpu\python.exe C:/Users/James/PycharmProjects/MLProject/Models.py
Using TensorFlow backend.
C:\Users\James\Anaconda3\envs\keras-gpu\lib\site-packages\sklearn\externals\six.py:31: DeprecationWarning: The module is deprecated in version 0.21 and will be removed in version 0.23 since we've dropped support for Python 2.7. Please rely on the official version of six (https://pypi.org/project/six/).
  "(https://pypi.org/project/six/).", DeprecationWarning)
The number of majority class instances in the training set is 140529
The number of minority class instances in the training set is 12367
The total is 152896
The number of majority class instances in the testing set is 93576
The number of minority class instances in the testing set is 6186
The total is 99762
After resampling
The number of majority class instances in the training set is 50000
The number of minority class instances in the training set is 20000
The total is 70000
The number of majority class instances in the validation set is 9986
The number of minority class instances in the validation set is 4014
The total is 14000

The accuracy for the model Decision Tree - Gini criterion is 85.04%
The minority class recall for the model Decision Tree - Gini criterion is 67.21%
The minority class f1 score for the model Decision Tree - Gini criterion is 72.04%
The minority class precision for the model Decision Tree - Gini criterion is 77.62%
The confusion matrix is given by: 
[[9208  778]
 [1316 2698]]

The accuracy for the model Decision Tree - Entropy criterion is 86.54%
The minority class recall for the model Decision Tree - Entropy criterion is 75.14%
The minority class f1 score for the model Decision Tree - Entropy criterion is 76.20%
The minority class precision for the model Decision Tree - Entropy criterion is 77.29%
The confusion matrix is given by: 
[[9100  886]
 [ 998 3016]]

The accuracy for the model Decision Tree - Gini criterion variant is 86.92%
The minority class recall for the model Decision Tree - Gini criterion variant is 81.54%
The minority class f1 score for the model Decision Tree - Gini criterion variant is 78.14%
The minority class precision for the model Decision Tree - Gini criterion variant is 75.02%
The confusion matrix is given by: 
[[8896 1090]
 [ 741 3273]]
In light of these results, the entropy decision tree will be chosen with max depth = 8, and minimum samples to split = 2

Printing the results per fold for the Decision Tree - Entropy criterion model

Fold 1: Accuracy  85.51%
Fold 1: Recall  65.80%
Fold 1: Precision  79.95%
Fold 1: ROC area  90.34%

Fold 2: Accuracy  84.77%
Fold 2: Recall  66.00%
Fold 2: Precision  77.37%
Fold 2: ROC area  89.98%

Fold 3: Accuracy  84.56%
Fold 3: Recall  65.75%
Fold 3: Precision  76.86%
Fold 3: ROC area  90.20%

Fold 4: Accuracy  85.07%
Fold 4: Recall  67.30%
Fold 4: Precision  77.49%
Fold 4: ROC area  90.40%

Fold 5: Accuracy  85.41%
Fold 5: Recall  67.90%
Fold 5: Precision  78.18%
Fold 5: ROC area  91.06%

Fold 6: Accuracy  84.37%
Fold 6: Recall  65.75%
Fold 6: Precision  76.28%
Fold 6: ROC area  89.62%

Fold 7: Accuracy  84.93%
Fold 7: Recall  69.65%
Fold 7: Precision  75.67%
Fold 7: ROC area  89.92%

Fold 8: Accuracy  84.97%
Fold 8: Recall  63.35%
Fold 8: Precision  79.89%
Fold 8: ROC area  89.90%

Fold 9: Accuracy  84.97%
Fold 9: Recall  67.85%
Fold 9: Precision  76.84%
Fold 9: ROC area  90.41%

Fold 10: Accuracy  84.96%
Fold 10: Recall  65.05%
Fold 10: Precision  78.61%
Fold 10: ROC area  89.90%

The accuracy for the model Decision Tree - Entropy criterion is 92.75%
The minority class recall for the model Decision Tree - Entropy criterion is 64.14%
The minority class f1 score for the model Decision Tree - Entropy criterion is 52.32%
The minority class precision for the model Decision Tree - Entropy criterion is 44.18%
The confusion matrix is given by: 
[[88563  5013]
 [ 2218  3968]]

The accuracy for the model Nearest Neighbours - Model 1  is 88.14%
The minority class recall for the model Nearest Neighbours - Model 1  is 83.23%
The minority class f1 score for the model Nearest Neighbours - Model 1  is 80.09%
The minority class precision for the model Nearest Neighbours - Model 1  is 77.18%
The confusion matrix is given by: 
[[8998  988]
 [ 673 3341]]

The accuracy for the model Nearest Neighbours - Model 2  is 84.94%
The minority class recall for the model Nearest Neighbours - Model 2  is 71.25%
The minority class f1 score for the model Nearest Neighbours - Model 2  is 73.07%
The minority class precision for the model Nearest Neighbours - Model 2  is 74.99%
The confusion matrix is given by: 
[[9032  954]
 [1154 2860]]

The accuracy for the model Nearest Neighbours - Model 3  is 88.15%
The minority class recall for the model Nearest Neighbours - Model 3  is 83.03%
The minority class f1 score for the model Nearest Neighbours - Model 3  is 80.07%
The minority class precision for the model Nearest Neighbours - Model 3  is 77.31%
The confusion matrix is given by: 
[[9008  978]
 [ 681 3333]]

The accuracy for the model Nearest Neighbours - Model 4  is 87.56%
The minority class recall for the model Nearest Neighbours - Model 4  is 83.13%
The minority class f1 score for the model Nearest Neighbours - Model 4  is 79.31%
The minority class precision for the model Nearest Neighbours - Model 4  is 75.82%
The confusion matrix is given by: 
[[8922 1064]
 [ 677 3337]]

The accuracy for the model Nearest Neighbours - Model 5  is 87.10%
The minority class recall for the model Nearest Neighbours - Model 5  is 81.46%
The minority class f1 score for the model Nearest Neighbours - Model 5  is 78.36%
The minority class precision for the model Nearest Neighbours - Model 5  is 75.48%
The confusion matrix is given by: 
[[8924 1062]
 [ 744 3270]]

As per the results, the first nearest neighbours model with parameters 5 neighbours at a weighted distance, will be chosen

Printing the results per fold for the Nearest Neighbours model

Fold 1: Accuracy  88.77%
Fold 1: Recall  84.85%
Fold 1: Precision  77.84%
Fold 1: ROC area  93.65%

Fold 2: Accuracy  88.03%
Fold 2: Recall  84.00%
Fold 2: Precision  76.43%
Fold 2: ROC area  93.03%

Fold 3: Accuracy  88.16%
Fold 3: Recall  83.70%
Fold 3: Precision  76.89%
Fold 3: ROC area  92.45%

Fold 4: Accuracy  87.76%
Fold 4: Recall  83.85%
Fold 4: Precision  75.85%
Fold 4: ROC area  92.98%

Fold 5: Accuracy  87.67%
Fold 5: Recall  84.40%
Fold 5: Precision  75.39%
Fold 5: ROC area  93.25%

Fold 6: Accuracy  88.47%
Fold 6: Recall  84.15%
Fold 6: Precision  77.45%
Fold 6: ROC area  93.23%

Fold 7: Accuracy  87.67%
Fold 7: Recall  82.90%
Fold 7: Precision  76.09%
Fold 7: ROC area  92.62%

Fold 8: Accuracy  88.94%
Fold 8: Recall  86.45%
Fold 8: Precision  77.46%
Fold 8: ROC area  93.75%

Fold 9: Accuracy  88.69%
Fold 9: Recall  85.15%
Fold 9: Precision  77.48%
Fold 9: ROC area  93.14%

Fold 10: Accuracy  88.37%
Fold 10: Recall  84.05%
Fold 10: Precision  77.25%
Fold 10: ROC area  93.10%

The accuracy for the model Nearest Neighbours is 90.71%
The minority class recall for the model Nearest Neighbours is 64.13%
The minority class f1 score for the model Nearest Neighbours is 46.11%
The minority class precision for the model Nearest Neighbours is 36.00%
The confusion matrix is given by: 
[[86524  7052]
 [ 2219  3967]]
2019-11-29 23:05:24.412700: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX AVX2
2019-11-29 23:05:24.414673: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library nvcuda.dll
Epoch 1/10
2019-11-29 23:05:24.430073: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1640] Found device 0 with properties: 
name: GeForce RTX 2080 major: 7 minor: 5 memoryClockRate(GHz): 1.8
pciBusID: 0000:01:00.0
2019-11-29 23:05:24.430280: I tensorflow/stream_executor/platform/default/dlopen_checker_stub.cc:25] GPU libraries are statically linked, skip dlopen check.
2019-11-29 23:05:24.430695: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1763] Adding visible gpu devices: 0
2019-11-29 23:05:24.817313: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1181] Device interconnect StreamExecutor with strength 1 edge matrix:
2019-11-29 23:05:24.817413: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1187]      0 
2019-11-29 23:05:24.817471: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1200] 0:   N 
2019-11-29 23:05:24.817997: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1326] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 6280 MB memory) -> physical GPU (device: 0, name: GeForce RTX 2080, pci bus id: 0000:01:00.0, compute capability: 7.5)

 1000/17500 [>.............................] - ETA: 16s - loss: 0.6989 - acc: 0.4000
17500/17500 [==============================] - 1s 60us/step - loss: 0.6869 - acc: 0.6256
Epoch 2/10

 1000/17500 [>.............................] - ETA: 0s - loss: 0.6707 - acc: 0.7460
17500/17500 [==============================] - 0s 3us/step - loss: 0.6546 - acc: 0.7429
Epoch 3/10

 1000/17500 [>.............................] - ETA: 0s - loss: 0.6366 - acc: 0.7450
17500/17500 [==============================] - 0s 2us/step - loss: 0.6138 - acc: 0.7490
Epoch 4/10

 1000/17500 [>.............................] - ETA: 0s - loss: 0.5960 - acc: 0.7400
17500/17500 [==============================] - 0s 2us/step - loss: 0.5729 - acc: 0.7552
Epoch 5/10

 1000/17500 [>.............................] - ETA: 0s - loss: 0.5436 - acc: 0.7600
17500/17500 [==============================] - 0s 2us/step - loss: 0.5351 - acc: 0.7659
Epoch 6/10

 1000/17500 [>.............................] - ETA: 0s - loss: 0.5223 - acc: 0.7800
17500/17500 [==============================] - 0s 2us/step - loss: 0.5055 - acc: 0.7745
Epoch 7/10

 1000/17500 [>.............................] - ETA: 0s - loss: 0.4778 - acc: 0.7870
17500/17500 [==============================] - 0s 3us/step - loss: 0.4811 - acc: 0.7845
Epoch 8/10

 1000/17500 [>.............................] - ETA: 0s - loss: 0.4787 - acc: 0.7720
17500/17500 [==============================] - 0s 2us/step - loss: 0.4581 - acc: 0.7921
Epoch 9/10

 1000/17500 [>.............................] - ETA: 0s - loss: 0.4527 - acc: 0.7940
17500/17500 [==============================] - 0s 2us/step - loss: 0.4337 - acc: 0.8007
Epoch 10/10

 1000/17500 [>.............................] - ETA: 0s - loss: 0.4281 - acc: 0.7930
17500/17500 [==============================] - 0s 2us/step - loss: 0.4082 - acc: 0.8095
Epoch 1/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.4026 - acc: 0.8060
8750/8750 [==============================] - 0s 3us/step - loss: 0.3934 - acc: 0.8171
Epoch 2/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.4058 - acc: 0.8090
8750/8750 [==============================] - 0s 3us/step - loss: 0.3845 - acc: 0.8213
Epoch 3/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3761 - acc: 0.8320
8750/8750 [==============================] - 0s 3us/step - loss: 0.3769 - acc: 0.8269
Epoch 4/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3786 - acc: 0.8260
8750/8750 [==============================] - 0s 3us/step - loss: 0.3719 - acc: 0.8298
Epoch 5/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3903 - acc: 0.8110
8750/8750 [==============================] - 0s 2us/step - loss: 0.3676 - acc: 0.8297
Epoch 6/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3578 - acc: 0.8320
8750/8750 [==============================] - 0s 2us/step - loss: 0.3652 - acc: 0.8323
Epoch 7/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3463 - acc: 0.8420
8750/8750 [==============================] - 0s 3us/step - loss: 0.3616 - acc: 0.8325
Epoch 8/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3324 - acc: 0.8520
8750/8750 [==============================] - 0s 2us/step - loss: 0.3597 - acc: 0.8335
Epoch 9/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3703 - acc: 0.8270
8750/8750 [==============================] - 0s 2us/step - loss: 0.3567 - acc: 0.8355
Epoch 10/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3388 - acc: 0.8530
8750/8750 [==============================] - 0s 3us/step - loss: 0.3547 - acc: 0.8351
Epoch 1/10

1000/7050 [===>..........................] - ETA: 0s - loss: 0.1204 - acc: 1.0000
7050/7050 [==============================] - 0s 3us/step - loss: 0.1128 - acc: 1.0000
Epoch 2/10

1000/7050 [===>..........................] - ETA: 0s - loss: 0.0930 - acc: 1.0000
7050/7050 [==============================] - 0s 3us/step - loss: 0.0804 - acc: 1.0000
Epoch 3/10

1000/7050 [===>..........................] - ETA: 0s - loss: 0.0636 - acc: 0.9990
7050/7050 [==============================] - 0s 3us/step - loss: 0.0541 - acc: 0.9999
Epoch 4/10

1000/7050 [===>..........................] - ETA: 0s - loss: 0.0429 - acc: 1.0000
7050/7050 [==============================] - 0s 3us/step - loss: 0.0382 - acc: 0.9991
Epoch 5/10

1000/7050 [===>..........................] - ETA: 0s - loss: 0.0317 - acc: 1.0000
7050/7050 [==============================] - 0s 3us/step - loss: 0.0287 - acc: 0.9997
Epoch 6/10

1000/7050 [===>..........................] - ETA: 0s - loss: 0.0248 - acc: 1.0000
7050/7050 [==============================] - 0s 2us/step - loss: 0.0227 - acc: 0.9997
Epoch 7/10

1000/7050 [===>..........................] - ETA: 0s - loss: 0.0191 - acc: 0.9990
7050/7050 [==============================] - 0s 3us/step - loss: 0.0184 - acc: 0.9996
Epoch 8/10

1000/7050 [===>..........................] - ETA: 0s - loss: 0.0162 - acc: 0.9990
7050/7050 [==============================] - 0s 3us/step - loss: 0.0154 - acc: 0.9997
Epoch 9/10

1000/7050 [===>..........................] - ETA: 0s - loss: 0.0128 - acc: 1.0000
7050/7050 [==============================] - 0s 3us/step - loss: 0.0132 - acc: 0.9999
Epoch 10/10

1000/7050 [===>..........................] - ETA: 0s - loss: 0.0122 - acc: 1.0000
7050/7050 [==============================] - 0s 3us/step - loss: 0.0114 - acc: 1.0000
Epoch 1/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.5667 - acc: 0.8300
8750/8750 [==============================] - 0s 2us/step - loss: 0.5620 - acc: 0.8297
Epoch 2/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.4584 - acc: 0.8430
8750/8750 [==============================] - 0s 3us/step - loss: 0.4624 - acc: 0.8327
Epoch 3/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3493 - acc: 0.8610
8750/8750 [==============================] - 0s 3us/step - loss: 0.3977 - acc: 0.8333
Epoch 4/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3834 - acc: 0.8380
8750/8750 [==============================] - 0s 3us/step - loss: 0.3723 - acc: 0.8362
Epoch 5/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3575 - acc: 0.8440
8750/8750 [==============================] - 0s 3us/step - loss: 0.3624 - acc: 0.8361
Epoch 6/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3606 - acc: 0.8450
8750/8750 [==============================] - 0s 3us/step - loss: 0.3570 - acc: 0.8390
Epoch 7/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3946 - acc: 0.8100
8750/8750 [==============================] - 0s 2us/step - loss: 0.3536 - acc: 0.8394
Epoch 8/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3315 - acc: 0.8450
8750/8750 [==============================] - 0s 2us/step - loss: 0.3518 - acc: 0.8409
Epoch 9/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3513 - acc: 0.8350
8750/8750 [==============================] - 0s 2us/step - loss: 0.3511 - acc: 0.8394
Epoch 10/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3556 - acc: 0.8260
8750/8750 [==============================] - 0s 2us/step - loss: 0.3497 - acc: 0.8401
Epoch 1/10

1000/7228 [===>..........................] - ETA: 0s - loss: 0.1230 - acc: 1.0000
7228/7228 [==============================] - 0s 3us/step - loss: 0.1143 - acc: 1.0000
Epoch 2/10

1000/7228 [===>..........................] - ETA: 0s - loss: 0.0981 - acc: 1.0000
7228/7228 [==============================] - 0s 3us/step - loss: 0.0939 - acc: 1.0000
Epoch 3/10

1000/7228 [===>..........................] - ETA: 0s - loss: 0.0865 - acc: 1.0000
7228/7228 [==============================] - 0s 2us/step - loss: 0.0729 - acc: 1.0000
Epoch 4/10

1000/7228 [===>..........................] - ETA: 0s - loss: 0.0607 - acc: 1.0000
7228/7228 [==============================] - 0s 3us/step - loss: 0.0572 - acc: 0.9992
Epoch 5/10

1000/7228 [===>..........................] - ETA: 0s - loss: 0.0511 - acc: 0.9990
7228/7228 [==============================] - 0s 3us/step - loss: 0.0463 - acc: 0.9996
Epoch 6/10

1000/7228 [===>..........................] - ETA: 0s - loss: 0.0351 - acc: 1.0000
7228/7228 [==============================] - 0s 3us/step - loss: 0.0381 - acc: 0.9997
Epoch 7/10

1000/7228 [===>..........................] - ETA: 0s - loss: 0.0344 - acc: 1.0000
7228/7228 [==============================] - 0s 2us/step - loss: 0.0321 - acc: 0.9997
Epoch 8/10

1000/7228 [===>..........................] - ETA: 0s - loss: 0.0311 - acc: 1.0000
7228/7228 [==============================] - 0s 3us/step - loss: 0.0275 - acc: 0.9997
Epoch 9/10

1000/7228 [===>..........................] - ETA: 0s - loss: 0.0245 - acc: 1.0000
7228/7228 [==============================] - 0s 2us/step - loss: 0.0240 - acc: 0.9997
Epoch 10/10

1000/7228 [===>..........................] - ETA: 0s - loss: 0.0237 - acc: 1.0000
7228/7228 [==============================] - 0s 3us/step - loss: 0.0211 - acc: 0.9997
Epoch 1/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.4828 - acc: 0.8440
8750/8750 [==============================] - 0s 2us/step - loss: 0.4770 - acc: 0.8351
Epoch 2/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.4540 - acc: 0.8330
8750/8750 [==============================] - 0s 3us/step - loss: 0.4264 - acc: 0.8369
Epoch 3/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.4531 - acc: 0.8240
8750/8750 [==============================] - 0s 3us/step - loss: 0.3851 - acc: 0.8387
Epoch 4/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.4062 - acc: 0.8310
8750/8750 [==============================] - 0s 3us/step - loss: 0.3628 - acc: 0.8413
Epoch 5/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3746 - acc: 0.8390
8750/8750 [==============================] - 0s 3us/step - loss: 0.3517 - acc: 0.8415
Epoch 6/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3544 - acc: 0.8340
8750/8750 [==============================] - 0s 3us/step - loss: 0.3464 - acc: 0.8414
Epoch 7/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3322 - acc: 0.8570
8750/8750 [==============================] - 0s 3us/step - loss: 0.3439 - acc: 0.8435
Epoch 8/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3244 - acc: 0.8570
8750/8750 [==============================] - 0s 3us/step - loss: 0.3418 - acc: 0.8416
Epoch 9/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3349 - acc: 0.8500
8750/8750 [==============================] - 0s 3us/step - loss: 0.3403 - acc: 0.8434
Epoch 10/10

1000/8750 [==>...........................] - ETA: 0s - loss: 0.3542 - acc: 0.8290
8750/8750 [==============================] - 0s 2us/step - loss: 0.3394 - acc: 0.8435
Epoch 1/10

1000/7268 [===>..........................] - ETA: 0s - loss: 0.1151 - acc: 1.0000
7268/7268 [==============================] - 0s 3us/step - loss: 0.1105 - acc: 1.0000
Epoch 2/10

1000/7268 [===>..........................] - ETA: 0s - loss: 0.1003 - acc: 1.0000
7268/7268 [==============================] - 0s 3us/step - loss: 0.0931 - acc: 1.0000
Epoch 3/10

1000/7268 [===>..........................] - ETA: 0s - loss: 0.0852 - acc: 1.0000
7268/7268 [==============================] - 0s 3us/step - loss: 0.0740 - acc: 0.9999
Epoch 4/10

1000/7268 [===>..........................] - ETA: 0s - loss: 0.0645 - acc: 0.9980
7268/7268 [==============================] - 0s 3us/step - loss: 0.0595 - acc: 0.9993
Epoch 5/10

1000/7268 [===>..........................] - ETA: 0s - loss: 0.0532 - acc: 1.0000
7268/7268 [==============================] - 0s 3us/step - loss: 0.0488 - acc: 0.9994
Epoch 6/10

1000/7268 [===>..........................] - ETA: 0s - loss: 0.0449 - acc: 0.9980
7268/7268 [==============================] - 0s 3us/step - loss: 0.0410 - acc: 0.9996
Epoch 7/10

1000/7268 [===>..........................] - ETA: 0s - loss: 0.0351 - acc: 1.0000
7268/7268 [==============================] - 0s 3us/step - loss: 0.0350 - acc: 1.0000
Epoch 8/10

1000/7268 [===>..........................] - ETA: 0s - loss: 0.0342 - acc: 1.0000
7268/7268 [==============================] - 0s 3us/step - loss: 0.0304 - acc: 1.0000
Epoch 9/10

1000/7268 [===>..........................] - ETA: 0s - loss: 0.0279 - acc: 1.0000
7268/7268 [==============================] - 0s 3us/step - loss: 0.0267 - acc: 1.0000
Epoch 10/10

1000/7268 [===>..........................] - ETA: 0s - loss: 0.0238 - acc: 1.0000
7268/7268 [==============================] - 0s 3us/step - loss: 0.0237 - acc: 1.0000

The accuracy for the model Semi-Supervised Neural Network is 90.29%
The minority class recall for the model Semi-Supervised Neural Network is 73.50%
The minority class f1 score for the model Semi-Supervised Neural Network is 48.42%
The minority class precision for the model Semi-Supervised Neural Network is 36.10%
The confusion matrix is given by: 
[[85526  8050]
 [ 1639  4547]]

The accuracy for the model Rule Classifier - Skope Rules - Model 1 is 81.28%
The minority class recall for the model Rule Classifier - Skope Rules - Model 1 is 67.91%
The minority class f1 score for the model Rule Classifier - Skope Rules - Model 1 is 67.53%
The minority class precision for the model Rule Classifier - Skope Rules - Model 1 is 67.16%
The confusion matrix is given by: 
[[8653 1333]
 [1288 2726]]

The accuracy for the model Rule Classifier - Skope Rules - Model 2 is 81.28%
The minority class recall for the model Rule Classifier - Skope Rules - Model 2 is 67.91%
The minority class f1 score for the model Rule Classifier - Skope Rules - Model 2 is 67.53%
The minority class precision for the model Rule Classifier - Skope Rules - Model 2 is 67.16%
The confusion matrix is given by: 
[[8653 1333]
 [1288 2726]]

The accuracy for the model Rule Classifier - Skope Rules - Model 3 is 80.96%
The minority class recall for the model Rule Classifier - Skope Rules - Model 3 is 70.48%
The minority class f1 score for the model Rule Classifier - Skope Rules - Model 3 is 67.98%
The minority class precision for the model Rule Classifier - Skope Rules - Model 3 is 65.65%
The confusion matrix is given by: 
[[8506 1480]
 [1185 2829]]

As per the results, Skope Rules model 1 will be chosen for training

Printing the results per fold for the Skope Rules model

Fold 1: Accuracy  76.73%
Fold 1: Recall  79.43%
Fold 1: Precision  57.29%
Fold 1: ROC area  82.46%

Fold 2: Accuracy  77.94%
Fold 2: Recall  78.43%
Fold 2: Precision  58.08%
Fold 2: ROC area  82.47%

Fold 3: Accuracy  81.27%
Fold 3: Recall  66.72%
Fold 3: Precision  66.34%
Fold 3: ROC area  78.26%

Fold 4: Accuracy  79.93%
Fold 4: Recall  71.24%
Fold 4: Precision  64.73%
Fold 4: ROC area  79.45%

Fold 5: Accuracy  80.96%
Fold 5: Recall  71.29%
Fold 5: Precision  64.91%
Fold 5: ROC area  79.91%

Fold 6: Accuracy  80.04%
Fold 6: Recall  67.02%
Fold 6: Precision  64.47%
Fold 6: ROC area  77.68%

Fold 7: Accuracy  76.16%
Fold 7: Recall  77.32%
Fold 7: Precision  55.72%
Fold 7: ROC area  80.40%

Fold 8: Accuracy  78.57%
Fold 8: Recall  76.89%
Fold 8: Precision  59.23%
Fold 8: ROC area  81.48%

Fold 9: Accuracy  80.73%
Fold 9: Recall  71.84%
Fold 9: Precision  64.98%
Fold 9: ROC area  80.32%

Fold 10: Accuracy  80.66%
Fold 10: Recall  68.45%
Fold 10: Precision  65.38%
Fold 10: ROC area  78.54%

The accuracy for the model Skope Rules is 87.51%
The minority class recall for the model Skope Rules is 70.01%
The minority class f1 score for the model Skope Rules is 41.01%
The minority class precision for the model Skope Rules is 29.00%
The confusion matrix is given by: 
[[82973 10603]
 [ 1855  4331]]
('Age > 0.33888889849185944 and OccupationCode > 0.3255113214254379', (0.7580615562888956, 0.47370105288521813, 6))
('Age > 0.33888889849185944 and CapitalGains <= 0.07073500007390976 and OccupationCode > 0.19287504255771637', (0.6329532619806674, 0.5339787787214668, 4))
('Age > 0.32777778804302216 and OccupationCode > 0.39162690937519073', (0.7588271732397275, 0.45706112803930105, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.39162690937519073', (0.7659686414816373, 0.453125, 1))
('Age > 0.32777778804302216 and OccupationCode > 0.21098510175943375 and StockDividends <= 0.0008749999979045242', (0.5995007132667618, 0.4206180407856875, 1))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.000885000015841797', (0.5974256963634389, 0.4192329790707881, 3))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.0008749999979045242', (0.5956894451888606, 0.4171034331982876, 2))
('Age > 0.33888889849185944 and OccupationCode > 0.21098510175943375 and StockDividends <= 0.000885000015841797', (0.6018756895917616, 0.4099448897795591, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.000885000015841797', (0.6030266678970194, 0.40836093232518905, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.22294022142887115 and StockDividends <= 0.000885000015841797', (0.6089411764705882, 0.40500782472613456, 1))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.0004599999956553802', (0.5918669439679892, 0.39968084595135406, 3))
('Age > 0.32777778804302216 and OccupationCode > 0.21098510175943375 and StockDividends <= 0.00040499999886378646', (0.5915597694738799, 0.39949780288763337, 1))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.00040499999886378646', (0.5923649434684248, 0.398571205946033, 2))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.0004749999934574589', (0.5902946292694915, 0.39861155155919087, 3))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.0004599999956553802', (0.5992934886385335, 0.3931233168409845, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.00043000000005122274', (0.5991387559808612, 0.3928100884622624, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.0004749999934574589', (0.6023324729949335, 0.3913178487144454, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.0004950000147800893', (0.5964176829268293, 0.39103004559935034, 1))
('Age > 0.32777778804302216 and OccupationCode > 0.21098510175943375 and StockDividends <= 0.000295000005280599', (0.5925856051315913, 0.3917435769518583, 1))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.000295000005280599', (0.5897159018229451, 0.3929756568440101, 3))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.0002649999951245263', (0.5899552435196991, 0.39282660068316616, 7))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.00040499999886378646', (0.5957878498498979, 0.38950045896761454, 2))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.000415000002249144', (0.5974527209571594, 0.38736315295589613, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.000295000005280599', (0.5976819053900045, 0.38473990591235524, 8))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.0002649999951245263', (0.5962753113776234, 0.38385154152613277, 15))
('Age > 0.33888889849185944 and OccupationCode > 0.21098510175943375 and StockDividends <= 0.0002649999951245263', (0.5970410492825486, 0.3823337276354584, 3))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.00014500000543193892', (0.5865964983757355, 0.381968603773535, 2))
('Age > 0.32777778804302216 and OccupationCode > 0.21098510175943375 and StockDividends <= 0.00017500000103609636', (0.5873259721555449, 0.3813828792318723, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.00017500000103609636', (0.5953565637781099, 0.3759515134143391, 4))
('Age > 0.32777778804302216 and OccupationCode > 0.22294022142887115 and StockDividends <= 0.00014500000543193892', (0.5950585433617781, 0.37507036087310025, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.21098510175943375 and StockDividends <= 0.00014500000543193892', (0.5945683991585221, 0.3724529726368242, 3))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.00014500000543193892', (0.5929806709304836, 0.37282727881586203, 14))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends <= 0.0001249999986612238', (0.5918854415274463, 0.3705409948328457, 1))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends > 0.00014500000543193892', (0.8193639123411108, 0.29256187530643324, 2))
('Age > 0.32777778804302216 and OccupationCode > 0.21098510175943375 and StockDividends > 0.00017500000103609636', (0.824290780141844, 0.2898559760583578, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends > 0.0001249999986612238', (0.824477506199079, 0.28979642657037913, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends > 0.00014500000543193892', (0.8262629633200241, 0.28761234425289445, 14))
('Age > 0.33888889849185944 and OccupationCode > 0.21098510175943375 and StockDividends > 0.00014500000543193892', (0.8254787195334413, 0.2873093678647693, 3))
('Age > 0.32777778804302216 and OccupationCode > 0.22294022142887115 and StockDividends > 0.00014500000543193892', (0.8254912565350639, 0.2863843892676215, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends > 0.00017500000103609636', (0.8253246096573672, 0.28455035006971263, 4))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends > 0.0002649999951245263', (0.8282374973249376, 0.2804031654503101, 7))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends > 0.000295000005280599', (0.8267373007584414, 0.27995157575131413, 3))
('Age > 0.32777778804302216 and OccupationCode > 0.21098510175943375 and StockDividends > 0.000295000005280599', (0.8292818704768974, 0.2786854577201297, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends > 0.0002649999951245263', (0.8301220038972548, 0.27708697765546, 15))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends > 0.000295000005280599', (0.830141113538927, 0.2765708261872993, 8))
('Age > 0.33888889849185944 and OccupationCode > 0.21098510175943375 and StockDividends > 0.0002649999951245263', (0.8312231871560821, 0.2751524150756574, 3))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends > 0.00040499999886378646', (0.8305568163297989, 0.2743689827548106, 2))
('Age > 0.32777778804302216 and OccupationCode > 0.21098510175943375 and StockDividends > 0.00040499999886378646', (0.8302604067667744, 0.27419962335216574, 1))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends > 0.0004599999956553802', (0.829527998679604, 0.2737504752331793, 3))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends > 0.0004749999934574589', (0.8304277039185107, 0.2734034341741089, 3))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends > 0.000415000002249144', (0.834739263803681, 0.27238035658429777, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends > 0.00040499999886378646', (0.8316926385026455, 0.2724605113304296, 2))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends > 0.0004950000147800893', (0.8380621501640609, 0.27122243737897433, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends > 0.0004599999956553802', (0.832786252172234, 0.2701196217197971, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends > 0.00043000000005122274', (0.8274661508704062, 0.2683982683982684, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends > 0.0004749999934574589', (0.8328155339805825, 0.26636442677928207, 1))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends > 0.0008749999979045242', (0.8410577845232624, 0.2560839854196901, 2))
('Age > 0.32777778804302216 and OccupationCode > 0.19287504255771637 and StockDividends > 0.000885000015841797', (0.8405542136571218, 0.25488125600654854, 3))
('Age > 0.33888889849185944 and OccupationCode > 0.21098510175943375 and StockDividends > 0.000885000015841797', (0.8403501458941226, 0.25250501002004005, 1))
('Age > 0.32777778804302216 and OccupationCode > 0.21098510175943375 and StockDividends > 0.0008749999979045242', (0.8378995433789954, 0.2525334667834355, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.19287504255771637 and StockDividends > 0.000885000015841797', (0.8381430363864492, 0.2504530400549897, 1))
('Age > 0.33888889849185944 and OccupationCode > 0.22294022142887115 and StockDividends > 0.000885000015841797', (0.8454095893356267, 0.24613458528951487, 1))
('Age > 0.32777778804302216 and OccupationCode <= 0.39162690937519073 and OccupationCode > 0.19287504255771637', (0.537722695584818, 0.2158447857720291, 1))
('Age > 0.33888889849185944 and OccupationCode <= 0.3255113214254379 and OccupationCode > 0.17949891835451126', (0.5137900355871886, 0.21690140845070421, 1))
('Age > 0.33888889849185944 and OccupationCode <= 0.39162690937519073 and OccupationCode > 0.19287504255771637', (0.5430771699021338, 0.21071962151394422, 1))

The accuracy for the model Ada Boost Classifier - Model 1 is 85.93%
The minority class recall for the model Ada Boost Classifier - Model 1 is 68.96%
The minority class f1 score for the model Ada Boost Classifier - Model 1 is 73.75%
The minority class precision for the model Ada Boost Classifier - Model 1 is 79.27%
The confusion matrix is given by: 
[[9262  724]
 [1246 2768]]

The accuracy for the model Ada Boost Classifier - Model 2 is 85.44%
The minority class recall for the model Ada Boost Classifier - Model 2 is 67.66%
The minority class f1 score for the model Ada Boost Classifier - Model 2 is 72.72%
The minority class precision for the model Ada Boost Classifier - Model 2 is 78.59%
The confusion matrix is given by: 
[[9246  740]
 [1298 2716]]

The accuracy for the model Ada Boost Classifier - Model 3 is 85.61%
The minority class recall for the model Ada Boost Classifier - Model 3 is 68.98%
The minority class f1 score for the model Ada Boost Classifier - Model 3 is 73.32%
The minority class precision for the model Ada Boost Classifier - Model 3 is 78.24%
The confusion matrix is given by: 
[[9216  770]
 [1245 2769]]

As per the results, Model 1 will be chosen for training and testing

Printing the results per fold for the Ada Boost Classifier model

Fold 1: Accuracy  85.84%
Fold 1: Recall  69.25%
Fold 1: Precision  78.65%
Fold 1: ROC area  91.91%

Fold 2: Accuracy  85.59%
Fold 2: Recall  68.80%
Fold 2: Precision  78.14%
Fold 2: ROC area  91.82%

Fold 3: Accuracy  85.49%
Fold 3: Recall  68.55%
Fold 3: Precision  77.99%
Fold 3: ROC area  91.69%

Fold 4: Accuracy  85.56%
Fold 4: Recall  68.80%
Fold 4: Precision  78.05%
Fold 4: ROC area  91.56%

Fold 5: Accuracy  85.89%
Fold 5: Recall  71.15%
Fold 5: Precision  77.59%
Fold 5: ROC area  92.28%

Fold 6: Accuracy  85.10%
Fold 6: Recall  67.80%
Fold 6: Precision  77.26%
Fold 6: ROC area  91.00%

Fold 7: Accuracy  85.33%
Fold 7: Recall  69.05%
Fold 7: Precision  77.19%
Fold 7: ROC area  91.48%

Fold 8: Accuracy  85.14%
Fold 8: Recall  67.30%
Fold 8: Precision  77.71%
Fold 8: ROC area  91.46%

Fold 9: Accuracy  85.56%
Fold 9: Recall  69.70%
Fold 9: Precision  77.49%
Fold 9: ROC area  91.80%

Fold 10: Accuracy  85.61%
Fold 10: Recall  68.70%
Fold 10: Precision  78.29%
Fold 10: ROC area  91.61%

The accuracy for the model Ada Boost Classifier is 92.58%
The minority class recall for the model Ada Boost Classifier is 69.29%
The minority class f1 score for the model Ada Boost Classifier is 53.67%
The minority class precision for the model Ada Boost Classifier is 43.80%
The confusion matrix is given by: 
[[88076  5500]
 [ 1900  4286]]

Printing stats for models Decision Tree and Nearest Neighbours
The Wilcoxon statistic for accuracy is given by 0.0
We reject the null hypothesis that the difference in accuracy of the models is not significantly different
The Wilcoxon statistic for recall is given by 0.0
We reject the null hypothesis that the difference in recall of the models is not significantly different

Printing stats for models Decision Tree and Skope Rules
The Wilcoxon statistic for accuracy is given by 0.0
We reject the null hypothesis that the difference in accuracy of the models is not significantly different
The Wilcoxon statistic for recall is given by 0.0
We reject the null hypothesis that the difference in recall of the models is not significantly different

Printing stats for models Decision Tree and Ada Boost Classifier
The Wilcoxon statistic for accuracy is given by 0.0
We reject the null hypothesis that the difference in accuracy of the models is not significantly different
The Wilcoxon statistic for recall is given by 1.0
We reject the null hypothesis that the difference in recall of the models is not significantly different

Printing stats for models Nearest Neighbours and Skope Rules
The Wilcoxon statistic for accuracy is given by 0.0
We reject the null hypothesis that the difference in accuracy of the models is not significantly different
The Wilcoxon statistic for recall is given by 0.0
We reject the null hypothesis that the difference in recall of the models is not significantly different

Printing stats for models Nearest Neighbours and Ada Boost Classifier
The Wilcoxon statistic for accuracy is given by 0.0
We reject the null hypothesis that the difference in accuracy of the models is not significantly different
The Wilcoxon statistic for recall is given by 0.0
We reject the null hypothesis that the difference in recall of the models is not significantly different

Printing stats for models Skope Rule and Ada Boost Classifier
The Wilcoxon statistic for accuracy is given by 0.0
We reject the null hypothesis that the difference in accuracy of the models is not significantly different
The Wilcoxon statistic for recall is given by 9.0
The recall of the models belongs is not significantly different

The accuracy for the model Overall Predictions is 92.37%
The minority class recall for the model Overall Predictions is 69.25%
The minority class f1 score for the model Overall Predictions is 52.97%
The minority class precision for the model Overall Predictions is 42.88%
The confusion matrix is given by: 
[[87870  5706]
 [ 1902  4284]]
The total running time of this script is 500.3 seconds

Process finished with exit code 0