In [1]:
import pandas as pd
import numpy as np
In [2]:
X = pd.read_csv('xtraining_SC_GGP_AXA_FR.csv')
y = pd.read_csv('ytraining_SC_GGP_AXA_FR.csv')
Xscore = pd.read_csv('xscoring_SC_GGP_AXA_FR.csv')
In [3]:
y.describe()
Out[3]:
target
count 1000.000000
mean 0.170000
std 0.375821
min 0.000000
25% 0.000000
50% 0.000000
75% 0.000000
max 1.000000
In [4]:
X.describe()
Out[4]:
age surface montant_assure nb_sinistres_depuis_ouverture cotisation_n-1 cotisation_n duree_contrat duree_dernier_sinistre duree_zero_sinistre depa_01 ... cate_Artisans, cate_Cadres cate_Employés cate_Inactifs cate_Ouvriers type_Appartement type_Maison stat_Locataire stat_Occupant stat_Propriétaire
count 1000.000000 1000.000000 1000.000000 1000.000000 1000.000000 1000.000000 1000.000000 1000.000000 1000.000000 1000.000000 ... 1000.000000 1000.000000 1000.000000 1000.000000 1000.000000 1000.00000 1000.00000 1000.000000 1000.000000 1000.000000
mean 42.198000 86.863400 22461.069000 1.492000 272.762170 286.046570 4883.704000 40.023000 3444.616000 0.009000 ... 0.174000 0.162000 0.341000 0.067000 0.125000 0.63400 0.36600 0.586000 0.028000 0.386000
std 14.710853 28.498626 9076.273609 0.954383 76.378759 82.097481 2716.954916 41.793913 2826.615494 0.094488 ... 0.379299 0.368635 0.474283 0.250147 0.330884 0.48195 0.48195 0.492795 0.165055 0.487074
min 15.000000 31.270000 4478.000000 0.000000 175.380000 181.360000 49.000000 0.000000 0.000000 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.00000 0.00000 0.000000 0.000000 0.000000
25% 31.000000 66.235000 15574.500000 1.000000 211.842500 220.230000 2637.750000 1.000000 565.000000 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.00000 0.00000 0.000000 0.000000 0.000000
50% 42.000000 80.605000 21201.000000 1.000000 227.525000 240.705000 4819.000000 29.000000 3111.500000 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 1.00000 0.00000 1.000000 0.000000 0.000000
75% 52.000000 104.842500 27597.000000 2.000000 360.062500 372.647500 7353.250000 68.000000 6037.000000 0.000000 ... 0.000000 0.000000 1.000000 0.000000 0.000000 1.00000 1.00000 1.000000 0.000000 1.000000
max 94.000000 179.780000 55852.000000 4.000000 440.870000 510.490000 9526.000000 202.000000 9429.000000 1.000000 ... 1.000000 1.000000 1.000000 1.000000 1.000000 1.00000 1.00000 1.000000 1.000000 1.000000

8 rows × 121 columns

In [5]:
Xscore.describe()
Out[5]:
age surface montant_assure nb_sinistres_depuis_ouverture cotisation_n-1 cotisation_n duree_contrat duree_dernier_sinistre duree_zero_sinistre depa_01 ... cate_Artisans, cate_Cadres cate_Employés cate_Inactifs cate_Ouvriers type_Appartement type_Maison stat_Locataire stat_Occupant stat_Propriétaire
count 300.00000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 ... 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000
mean 42.59000 86.704767 22894.856667 1.466667 274.989200 288.086200 4810.156667 34.820000 3174.016667 0.010000 ... 0.156667 0.186667 0.370000 0.023333 0.116667 0.626667 0.373333 0.560000 0.016667 0.423333
std 14.62949 30.453970 9374.614330 0.992727 80.104622 86.567491 2673.566074 39.671826 2745.057079 0.099665 ... 0.364094 0.390295 0.483611 0.151212 0.321559 0.484498 0.484498 0.497216 0.128233 0.494913
min 15.00000 29.360000 7217.000000 0.000000 172.470000 180.700000 88.000000 0.000000 0.000000 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
25% 31.00000 63.672500 15281.750000 1.000000 212.580000 221.105000 2521.000000 1.000000 472.500000 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
50% 43.00000 78.530000 20810.500000 1.000000 227.590000 240.395000 5101.000000 19.500000 2566.500000 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 1.000000 0.000000 0.000000
75% 54.00000 107.357500 28494.750000 2.000000 364.185000 370.237500 7099.500000 62.000000 5547.250000 0.000000 ... 0.000000 0.000000 1.000000 0.000000 0.000000 1.000000 1.000000 1.000000 0.000000 1.000000
max 86.00000 169.560000 50948.000000 5.000000 450.560000 599.500000 9530.000000 189.000000 9019.000000 1.000000 ... 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000

8 rows × 121 columns

In [6]:
from sklearn.preprocessing import MinMaxScaler

scaler = MinMaxScaler()
X = scaler.fit_transform(X)
Xscore = scaler.transform(Xscore)
In [7]:
pd.DataFrame(X).describe()
Out[7]:
0 1 2 3 4 5 6 7 8 9 ... 111 112 113 114 115 116 117 118 119 120
count 1000.000000 1000.000000 1000.000000 1000.000000 1000.000000 1000.000000 1000.000000 1000.000000 1000.000000 1000.000000 ... 1000.000000 1000.000000 1000.000000 1000.000000 1000.000000 1000.00000 1000.00000 1000.000000 1000.000000 1000.000000
mean 0.344278 0.374341 0.350042 0.373000 0.366802 0.318071 0.510151 0.198134 0.365321 0.009000 ... 0.174000 0.162000 0.341000 0.067000 0.125000 0.63400 0.36600 0.586000 0.028000 0.386000
std 0.186213 0.191897 0.176671 0.238596 0.287690 0.249438 0.286689 0.206901 0.299779 0.094488 ... 0.379299 0.368635 0.474283 0.250147 0.330884 0.48195 0.48195 0.492795 0.165055 0.487074
min 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.00000 0.00000 0.000000 0.000000 0.000000
25% 0.202532 0.235439 0.215994 0.250000 0.137340 0.118099 0.273161 0.004950 0.059922 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.00000 0.00000 0.000000 0.000000 0.000000
50% 0.341772 0.332200 0.325515 0.250000 0.196410 0.180309 0.503324 0.143564 0.329993 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 1.00000 0.00000 1.000000 0.000000 0.000000
75% 0.468354 0.495404 0.450014 0.500000 0.695629 0.581191 0.770734 0.336634 0.640259 0.000000 ... 0.000000 0.000000 1.000000 0.000000 0.000000 1.00000 1.00000 1.000000 0.000000 1.000000
max 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 ... 1.000000 1.000000 1.000000 1.000000 1.000000 1.00000 1.00000 1.000000 1.000000 1.000000

8 rows × 121 columns

In [8]:
pd.DataFrame(Xscore).describe()
Out[8]:
0 1 2 3 4 5 6 7 8 9 ... 111 112 113 114 115 116 117 118 119 120
count 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 ... 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000 300.000000
mean 0.349241 0.373273 0.358486 0.366667 0.375190 0.324268 0.502391 0.172376 0.336623 0.010000 ... 0.156667 0.186667 0.370000 0.023333 0.116667 0.626667 0.373333 0.560000 0.016667 0.423333
std 0.185183 0.205063 0.182478 0.248182 0.301724 0.263019 0.282111 0.196395 0.291129 0.099665 ... 0.364094 0.390295 0.483611 0.151212 0.321559 0.484498 0.484498 0.497216 0.128233 0.494913
min 0.000000 -0.012861 0.053315 0.000000 -0.010961 -0.002005 0.004115 0.000000 0.000000 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
25% 0.202532 0.218184 0.210296 0.250000 0.140118 0.120758 0.260842 0.004950 0.050111 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
50% 0.354430 0.318228 0.317914 0.250000 0.196655 0.179367 0.533080 0.096535 0.272192 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 1.000000 0.000000 0.000000
75% 0.493671 0.512339 0.467488 0.500000 0.711157 0.573869 0.743959 0.306931 0.588318 0.000000 ... 0.000000 0.000000 1.000000 0.000000 0.000000 1.000000 1.000000 1.000000 0.000000 1.000000
max 0.898734 0.931183 0.904543 1.250000 1.036499 1.270440 1.000422 0.935644 0.956517 1.000000 ... 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000

8 rows × 121 columns

In [9]:
from sklearn.model_selection import train_test_split  

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
len(X_test)/len(X)
Out[9]:
0.2
In [10]:
pd.DataFrame(X_train).describe()
Out[10]:
0 1 2 3 4 5 6 7 8 9 ... 111 112 113 114 115 116 117 118 119 120
count 800.000000 800.000000 800.000000 800.000000 800.000000 800.000000 800.000000 800.000000 800.000000 800.000000 ... 800.000000 800.000000 800.000000 800.000000 800.000000 800.00000 800.00000 800.000000 800.000000 800.000000
mean 0.344747 0.373978 0.352949 0.368750 0.368131 0.318731 0.506860 0.197333 0.358045 0.006250 ... 0.172500 0.165000 0.342500 0.073750 0.122500 0.63375 0.36625 0.580000 0.027500 0.392500
std 0.185341 0.189740 0.180912 0.234917 0.286502 0.249066 0.288273 0.201852 0.298163 0.078859 ... 0.378051 0.371413 0.474843 0.261527 0.328068 0.48208 0.48208 0.493867 0.163637 0.488613
min 0.000000 0.010774 0.000000 0.000000 0.000000 0.000000 0.001688 0.000000 0.000000 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.00000 0.00000 0.000000 0.000000 0.000000
25% 0.202532 0.235439 0.215002 0.250000 0.138178 0.119444 0.266092 0.004950 0.057376 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.00000 0.00000 0.000000 0.000000 0.000000
50% 0.341772 0.335903 0.328045 0.250000 0.199066 0.180445 0.499156 0.153465 0.312971 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 1.00000 0.00000 1.000000 0.000000 0.000000
75% 0.468354 0.488570 0.460938 0.500000 0.694037 0.581207 0.768756 0.336634 0.627559 0.000000 ... 0.000000 0.000000 1.000000 0.000000 0.000000 1.00000 1.00000 1.000000 0.000000 1.000000
max 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 ... 1.000000 1.000000 1.000000 1.000000 1.000000 1.00000 1.00000 1.000000 1.000000 1.000000

8 rows × 121 columns

In [11]:
pd.DataFrame(X_test).describe()
Out[11]:
0 1 2 3 4 5 6 7 8 9 ... 111 112 113 114 115 116 117 118 119 120
count 200.000000 200.000000 200.000000 200.000000 200.000000 200.000000 200.000000 200.000000 200.000000 200.000000 ... 200.000000 200.000000 200.000000 200.000000 200.000000 200.000000 200.000000 200.000000 200.000000 200.000000
mean 0.342405 0.375793 0.338414 0.390000 0.361484 0.315430 0.523317 0.201337 0.394428 0.020000 ... 0.180000 0.150000 0.335000 0.040000 0.135000 0.635000 0.365000 0.610000 0.030000 0.360000
std 0.190125 0.200777 0.158461 0.252674 0.293059 0.251531 0.280589 0.226485 0.305180 0.140351 ... 0.385152 0.357967 0.473175 0.196451 0.342581 0.482638 0.482638 0.488974 0.171015 0.481205
min 0.000000 0.000000 0.053295 0.000000 0.016498 0.028742 0.000000 0.000000 0.000000 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
25% 0.215190 0.235506 0.231868 0.250000 0.129421 0.111909 0.287802 0.004950 0.077871 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
50% 0.341772 0.319541 0.310498 0.500000 0.185242 0.179564 0.532605 0.123762 0.390550 0.000000 ... 0.000000 0.000000 0.000000 0.000000 0.000000 1.000000 0.000000 1.000000 0.000000 0.000000
75% 0.443038 0.518955 0.424106 0.500000 0.707051 0.580143 0.772475 0.329208 0.671148 0.000000 ... 0.000000 0.000000 1.000000 0.000000 0.000000 1.000000 1.000000 1.000000 0.000000 1.000000
max 0.949367 0.982291 0.834294 1.000000 0.890956 0.831921 0.996307 0.930693 0.968607 1.000000 ... 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000 1.000000

8 rows × 121 columns

In [12]:
pd.DataFrame(y_train).describe()
Out[12]:
target
count 800.000000
mean 0.170000
std 0.375868
min 0.000000
25% 0.000000
50% 0.000000
75% 0.000000
max 1.000000
In [13]:
pd.DataFrame(y_test).describe()
Out[13]:
target
count 200.000000
mean 0.170000
std 0.376575
min 0.000000
25% 0.000000
50% 0.000000
75% 0.000000
max 1.000000
In [14]:
from imblearn.over_sampling import SMOTE

sm = SMOTE(random_state=42)
x_train_res, y_train_res = sm.fit_sample(X_train, y_train)

print(y_train.value_counts(), y_train_res.value_counts())

target
0.0       664
1.0       136
dtype: int64 target
1.0       664
0.0       664
dtype: int64
In [15]:
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import classification_report

model_RF = RandomForestClassifier(n_estimators=1000)
model_RF.fit(x_train_res , y_train_res)
y_pred = model_RF.predict(X_test)
target_names = ['non résilié', 'resilié'] # 1 résilié, 0 pas résilié
print(classification_report(y_test, y_pred,target_names=target_names))

              precision    recall  f1-score   support

 non résilié       0.93      0.92      0.92       166
     resilié       0.62      0.68      0.65        34

    accuracy                           0.88       200
   macro avg       0.78      0.80      0.79       200
weighted avg       0.88      0.88      0.88       200
In [16]:
from sklearn.ensemble import GradientBoostingClassifier

model_GB = GradientBoostingClassifier(n_estimators=1000)
model_GB.fit(x_train_res , y_train_res)
y_pred = model_GB.predict(X_test)
target_names = ['non résilié', 'resilié'] 
print(classification_report(y_test, y_pred,target_names=target_names))

              precision    recall  f1-score   support

 non résilié       0.94      0.91      0.92       166
     resilié       0.62      0.71      0.66        34

    accuracy                           0.88       200
   macro avg       0.78      0.81      0.79       200
weighted avg       0.88      0.88      0.88       200
In [17]:
from sklearn.ensemble import AdaBoostClassifier

model_AB = AdaBoostClassifier()
model_AB.fit(x_train_res , y_train_res)
y_pred = model_AB.predict(X_test)
target_names = ['non résilié', 'resilié'] 
print(classification_report(y_test, y_pred,target_names=target_names))

              precision    recall  f1-score   support

 non résilié       0.93      0.88      0.90       166
     resilié       0.53      0.68      0.60        34

    accuracy                           0.84       200
   macro avg       0.73      0.78      0.75       200
weighted avg       0.86      0.84      0.85       200
In [18]:
from bayes_opt import BayesianOptimization
from sklearn.metrics import f1_score

# Basic Example : https://github.com/fmfn/BayesianOptimization/blob/master/examples/basic-tour.ipynb

def black_box_function(n_estimators, learning_rate=0):
    model = AdaBoostClassifier(n_estimators = int(n_estimators),
                               learning_rate = 10**learning_rate,
                               random_state = 42)
    model.fit(x_train_res , y_train_res)
    y_pred = model.predict(X_test)
    return f1_score(y_test, y_pred)

pbounds = {'n_estimators':(10, 30), 'learning_rate':(-0.05, 0.05)}

optimizer = BayesianOptimization(
    f=black_box_function,
    pbounds=pbounds,
    verbose=2, # verbose = 1 prints only when a maximum is observed, verbose = 0 is silent
    random_state=1,
)

optimizer.maximize(
    init_points=10,
    n_iter=200,
)

print('max')

print(optimizer.max)

|   iter    |  target   | learni... | n_esti... |
-------------------------------------------------
|  1        |  0.7089   | -0.008298 |  24.41    |
|  2        |  0.6923   | -0.04999  |  16.05    |
|  3        |  0.7073   | -0.03532  |  11.85    |
|  4        |  0.6897   | -0.03137  |  16.91    |
|  5        |  0.6829   | -0.01032  |  20.78    |
|  6        |  0.72     | -0.008081 |  23.7     |
|  7        |  0.6753   | -0.02955  |  27.56    |
|  8        |  0.7105   | -0.04726  |  23.41    |
|  9        |  0.7273   | -0.00827  |  21.17    |
|  10       |  0.7229   | -0.03596  |  13.96    |
|  11       |  0.7179   | -0.007011 |  21.28    |
|  12       |  0.7073   |  0.01039  |  13.78    |
|  13       |  0.7317   | -0.03507  |  22.58    |
|  14       |  0.7013   |  0.01136  |  22.72    |
|  15       |  0.6977   | -0.01044  |  13.23    |
|  16       |  0.7105   | -0.05     |  22.49    |
|  17       |  0.6829   | -0.01468  |  21.17    |
|  18       |  0.72     |  0.01099  |  17.86    |
|  19       |  0.6923   |  0.00532  |  24.04    |
|  20       |  0.6966   | -0.03516  |  18.13    |
|  21       |  0.7073   |  0.01034  |  13.78    |
|  22       |  0.7273   | -0.03031  |  22.57    |
|  23       |  0.6829   | -0.01265  |  21.18    |
|  24       |  0.7317   | -0.03232  |  22.58    |
|  25       |  0.6835   | -0.03319  |  23.68    |
|  26       |  0.7229   | -0.03413  |  13.97    |
|  27       |  0.7297   | -0.02612  |  22.58    |
|  28       |  0.7013   |  0.000315 |  21.18    |
|  29       |  0.7073   | -0.02756  |  22.59    |
|  30       |  0.72     |  0.000387 |  23.7     |
|  31       |  0.6923   |  0.003145 |  23.71    |
|  32       |  0.7229   | -0.02833  |  13.96    |
|  33       |  0.7105   | -0.04397  |  22.58    |
|  34       |  0.6842   | -0.003856 |  19.76    |
|  35       |  0.7059   | -0.03591  |  22.57    |
|  36       |  0.7229   | -0.01857  |  14.29    |
|  37       |  0.72     | -0.00727  |  23.7     |
|  38       |  0.7368   | -0.02643  |  14.29    |
|  39       |  0.6923   | -0.01537  |  22.58    |
|  40       |  0.7229   | -0.02017  |  14.28    |
|  41       |  0.7143   | -0.02964  |  14.29    |
|  42       |  0.7143   | -0.03763  |  14.28    |
|  43       |  0.6977   | -0.02414  |  14.28    |
|  44       |  0.72     | -0.004892 |  23.7     |
|  45       |  0.725    | -0.006783 |  27.31    |
|  46       |  0.7143   | -0.0317   |  14.29    |
|  47       |  0.7059   | -0.03789  |  22.58    |
|  48       |  0.7229   | -0.02697  |  13.96    |
|  49       |  0.7273   | -0.02878  |  22.58    |
|  50       |  0.725    | -0.004436 |  27.31    |
|  51       |  0.7179   | -0.006697 |  21.28    |
|  52       |  0.7273   | -0.02933  |  22.59    |
|  53       |  0.7143   | -0.01339  |  14.28    |
|  54       |  0.6835   | -0.03413  |  23.39    |
|  55       |  0.6923   | -0.02074  |  29.96    |
|  56       |  0.7229   | -0.03278  |  13.96    |
|  57       |  0.6977   | -0.02405  |  14.29    |
|  58       |  0.7297   | -0.02543  |  22.58    |
|  59       |  0.7229   | -0.03681  |  13.97    |
|  60       |  0.6494   | -0.009551 |  27.31    |
|  61       |  0.6829   |  0.006223 |  16.56    |
|  62       |  0.6933   |  0.02765  |  13.92    |
|  63       |  0.7317   | -0.03269  |  22.59    |
|  64       |  0.6753   | -0.02335  |  22.58    |
|  65       |  0.7073   | -0.03869  |  13.97    |
|  66       |  0.7229   | -0.03417  |  13.96    |
|  67       |  0.7229   | -0.0304   |  13.96    |
|  68       |  0.7105   |  0.02672  |  23.46    |
|  69       |  0.7073   | -0.03992  |  13.97    |
|  70       |  0.7317   | -0.03472  |  22.58    |
|  71       |  0.6753   |  0.01585  |  19.81    |
|  72       |  0.7297   | -0.02563  |  22.57    |
|  73       |  0.7297   | -0.02604  |  22.58    |
|  74       |  0.7317   | -0.03239  |  22.57    |
|  75       |  0.7105   | -0.02205  |  22.57    |
|  76       |  0.7209   | -0.02868  |  19.62    |
|  77       |  0.7059   | -0.03574  |  22.57    |
|  78       |  0.725    | -0.002359 |  27.31    |
|  79       |  0.7273   | -0.02864  |  22.57    |
|  80       |  0.6753   |  0.03513  |  20.67    |
|  81       |  0.7143   | -0.02885  |  14.29    |
|  82       |  0.72     | -0.000691 |  23.7     |
|  83       |  0.7143   | -0.02838  |  14.28    |
|  84       |  0.7229   | -0.0215   |  14.27    |
|  85       |  0.725    | -0.001135 |  27.31    |
|  86       |  0.7273   | -0.02949  |  22.58    |
|  87       |  0.7143   | -0.01783  |  14.27    |
|  88       |  0.7143   | -0.01541  |  14.28    |
|  89       |  0.7229   | -0.03371  |  13.97    |
|  90       |  0.6667   |  0.03289  |  29.7     |
|  91       |  0.7013   | -0.004525 |  21.17    |
|  92       |  0.725    | -0.006537 |  27.3     |
|  93       |  0.7317   | -0.03363  |  22.58    |
|  94       |  0.72     |  0.01181  |  17.85    |
|  95       |  0.7273   | -0.02811  |  22.56    |
|  96       |  0.7179   |  0.03583  |  11.03    |
|  97       |  0.6988   | -0.04844  |  11.03    |
|  98       |  0.72     | -0.005421 |  23.69    |
|  99       |  0.725    | -0.003373 |  27.31    |
|  100      |  0.7229   | -0.02875  |  13.97    |
|  101      |  0.7317   | -0.0317   |  22.57    |
|  102      |  0.725    | -0.003994 |  27.3     |
|  103      |  0.7317   | -0.02606  |  13.97    |
|  104      |  0.7317   | -0.03203  |  22.56    |
|  105      |  0.6667   | -0.01108  |  23.7     |
|  106      |  0.7229   | -0.03034  |  13.97    |
|  107      |  0.7273   | -0.02879  |  22.56    |
|  108      |  0.725    |  0.001284 |  27.3     |
|  109      |  0.6977   | -0.02414  |  13.96    |
|  110      |  0.7229   | -0.02835  |  13.96    |
|  111      |  0.7317   | -0.02569  |  13.97    |
|  112      |  0.7059   | -0.03629  |  22.56    |
|  113      |  0.7317   | -0.03362  |  22.55    |
|  114      |  0.7297   | -0.0264   |  22.55    |
|  115      |  0.7105   | -0.02175  |  22.55    |
|  116      |  0.6753   |  0.02889  |  20.23    |
|  117      |  0.7105   | -0.03839  |  22.55    |
|  118      |  0.6494   | -0.00918  |  27.3     |
|  119      |  0.7027   |  0.02998  |  27.67    |
|  120      |  0.6818   | -0.01368  |  10.26    |
|  121      |  0.7229   | -0.02186  |  13.98    |
|  122      |  0.7273   | -0.02985  |  22.55    |
|  123      |  0.72     |  0.01326  |  17.85    |
|  124      |  0.7073   | -0.03763  |  12.63    |
|  125      |  0.7317   | -0.02649  |  13.98    |
|  126      |  0.725    |  0.000708 |  27.31    |
|  127      |  0.7073   | -0.04029  |  13.49    |
|  128      |  0.6818   | -0.0342   |  19.62    |
|  129      |  0.6977   | -0.02281  |  13.97    |
|  130      |  0.7342   | -0.04295  |  20.19    |
|  131      |  0.7342   | -0.04592  |  20.19    |
|  132      |  0.7342   | -0.0392   |  20.18    |
|  133      |  0.6842   |  0.002835 |  11.47    |
|  134      |  0.7105   | -0.022    |  22.56    |
|  135      |  0.7317   | -0.03169  |  22.55    |
|  136      |  0.7342   | -0.04505  |  20.19    |
|  137      |  0.6923   | -0.01419  |  22.83    |
|  138      |  0.6966   | -0.03732  |  18.91    |
|  139      |  0.6977   | -0.03507  |  20.18    |
|  140      |  0.7229   | -0.02949  |  13.97    |
|  141      |  0.7143   | -0.01192  |  14.29    |
|  142      |  0.7436   | -0.04955  |  20.19    |
|  143      |  0.7317   | -0.02579  |  13.98    |
|  144      |  0.7059   | -0.03712  |  22.55    |
|  145      |  0.7436   | -0.04984  |  20.19    |
|  146      |  0.7342   | -0.04167  |  20.18    |
|  147      |  0.7105   | -0.03834  |  22.59    |
|  148      |  0.7342   | -0.04779  |  20.2     |
|  149      |  0.6494   | -0.0114   |  27.3     |
|  150      |  0.7342   | -0.04754  |  20.18    |
|  151      |  0.7229   | -0.02047  |  11.35    |
|  152      |  0.7229   | -0.02718  |  13.98    |
|  153      |  0.7342   | -0.04512  |  20.18    |
|  154      |  0.7294   | -0.0198   |  19.62    |
|  155      |  0.7294   | -0.02184  |  19.61    |
|  156      |  0.7654   | -0.02304  |  19.61    |
|  157      |  0.7294   | -0.01931  |  19.61    |
|  158      |  0.6966   | -0.03566  |  18.6     |
|  159      |  0.7436   | -0.04827  |  20.2     |
|  160      |  0.6923   | -0.02572  |  19.61    |
|  161      |  0.7342   | -0.04471  |  20.18    |
|  162      |  0.7342   | -0.04209  |  20.18    |
|  163      |  0.6316   | -0.01572  |  24.41    |
|  164      |  0.7342   | -0.04472  |  20.17    |
|  165      |  0.6966   | -0.02401  |  19.62    |
|  166      |  0.7229   | -0.01921  |  13.98    |
|  167      |  0.7229   | -0.03052  |  13.98    |
|  168      |  0.7342   | -0.04705  |  20.2     |
|  169      |  0.7      |  0.04646  |  12.87    |
|  170      |  0.6923   | -0.02636  |  19.62    |
|  171      |  0.7436   | -0.04914  |  20.17    |
|  172      |  0.7342   | -0.04669  |  20.17    |
|  173      |  0.7342   | -0.04572  |  20.2     |
|  174      |  0.7273   | -0.04561  |  18.47    |
|  175      |  0.7848   |  0.04772  |  17.78    |
|  176      |  0.7848   |  0.04406  |  17.78    |
|  177      |  0.7848   |  0.04471  |  17.78    |
|  178      |  0.7317   |  0.03386  |  18.93    |
|  179      |  0.7632   |  0.04011  |  18.93    |
|  180      |  0.7848   |  0.04769  |  17.78    |
|  181      |  0.7317   |  0.03206  |  18.94    |
|  182      |  0.7532   |  0.04485  |  18.92    |
|  183      |  0.7632   |  0.042    |  18.93    |
|  184      |  0.72     |  0.01095  |  17.86    |
|  185      |  0.7143   | -0.04837  |  15.47    |
|  186      |  0.7848   |  0.0473   |  17.78    |
|  187      |  0.6829   | -0.0407   |  23.65    |
|  188      |  0.6923   | -0.01221  |  26.04    |
|  189      |  0.7532   |  0.04529  |  18.94    |
|  190      |  0.7848   |  0.04348  |  17.79    |
|  191      |  0.7      |  0.04061  |  12.46    |
|  192      |  0.7848   |  0.04312  |  17.79    |
|  193      |  0.7532   |  0.04991  |  18.93    |
|  194      |  0.7848   |  0.04101  |  17.79    |
|  195      |  0.7073   |  0.0367   |  17.79    |
|  196      |  0.7848   |  0.0435   |  17.78    |
|  197      |  0.6923   |  0.02793  |  24.72    |
|  198      |  0.7013   |  0.01218  |  22.56    |
|  199      |  0.7073   |  0.007552 |  13.13    |
|  200      |  0.6761   |  0.03671  |  28.28    |
|  201      |  0.7105   |  0.03921  |  18.93    |
|  202      |  0.7848   |  0.04867  |  17.78    |
|  203      |  0.7436   |  0.03745  |  18.93    |
|  204      |  0.7532   |  0.04703  |  18.93    |
|  205      |  0.7317   |  0.02761  |  18.92    |
|  206      |  0.7532   |  0.04951  |  18.93    |
|  207      |  0.7848   |  0.04445  |  17.77    |
|  208      |  0.7848   |  0.04138  |  17.79    |
|  209      |  0.6667   | -0.02362  |  23.7     |
|  210      |  0.7632   |  0.04247  |  18.93    |
=================================================
max
{'target': 0.7848101265822784, 'params': {'learning_rate': 0.047715766463314324, 'n_estimators': 17.779424609189796}}
In [19]:
model = AdaBoostClassifier(n_estimators = int(optimizer.max['params']['n_estimators']),
                            learning_rate = 10**optimizer.max['params']['learning_rate'],
                            random_state = 42)
model.fit(x_train_res , y_train_res)
y_pred = model.predict(X_test)
print("F1", f1_score(y_test, y_pred))
print(classification_report(y_test, y_pred,target_names=target_names))

F1 0.7848101265822784
              precision    recall  f1-score   support

 non résilié       0.98      0.92      0.95       166
     resilié       0.69      0.91      0.78        34

    accuracy                           0.92       200
   macro avg       0.83      0.91      0.87       200
weighted avg       0.93      0.92      0.92       200
In [20]:
yscore = model.predict(Xscore)
np.mean(yscore)
Out[20]:
0.2733333333333333
In [21]:
yDF = pd.DataFrame(data = yscore.astype(int), columns=['target'])
yDF.index+=1000
yDF.to_csv("result_schindler_hugo.csv")
In [ ]: