notebook.community

Edit and run 





In [1]:



    


import os
import pandas as pd
import math
import numpy as np
from sklearn.neural_network import MLPClassifier
from sklearn.preprocessing import StandardScaler



    











In [29]:



    


data = pd.read_csv('../kaggle_porto_seguro/train.csv')
data.head()



    


















    
Out[29]:










  
    

      
      id
      target
      ps_ind_01
      ps_ind_02_cat
      ps_ind_03
      ps_ind_04_cat
      ps_ind_05_cat
      ps_ind_06_bin
      ps_ind_07_bin
      ps_ind_08_bin
      ...
      ps_calc_11
      ps_calc_12
      ps_calc_13
      ps_calc_14
      ps_calc_15_bin
      ps_calc_16_bin
      ps_calc_17_bin
      ps_calc_18_bin
      ps_calc_19_bin
      ps_calc_20_bin
    

  
  
    

      0
      7
      0
      2
      2
      5
      1
      0
      0
      1
      0
      ...
      9
      1
      5
      8
      0
      1
      1
      0
      0
      1
    

    

      1
      9
      0
      1
      1
      7
      0
      0
      0
      0
      1
      ...
      3
      1
      1
      9
      0
      1
      1
      0
      1
      0
    

    

      2
      13
      0
      5
      4
      9
      1
      0
      0
      0
      1
      ...
      4
      2
      7
      7
      0
      1
      1
      0
      1
      0
    

    

      3
      16
      0
      0
      1
      2
      0
      0
      1
      0
      0
      ...
      2
      2
      4
      9
      0
      0
      0
      0
      0
      0
    

    

      4
      17
      0
      0
      2
      0
      1
      0
      1
      0
      0
      ...
      3
      1
      1
      3
      0
      0
      0
      1
      1
      0
    

  



5 rows × 59 columns



















In [30]:



    


pos_data = data[(data.target == 1)]
neg_data = data[(data.target == 0)]
pos_data.info()



    


















    






<class 'pandas.core.frame.DataFrame'>
Int64Index: 21694 entries, 9 to 595158
Data columns (total 59 columns):
id                21694 non-null int64
target            21694 non-null int64
ps_ind_01         21694 non-null int64
ps_ind_02_cat     21694 non-null int64
ps_ind_03         21694 non-null int64
ps_ind_04_cat     21694 non-null int64
ps_ind_05_cat     21694 non-null int64
ps_ind_06_bin     21694 non-null int64
ps_ind_07_bin     21694 non-null int64
ps_ind_08_bin     21694 non-null int64
ps_ind_09_bin     21694 non-null int64
ps_ind_10_bin     21694 non-null int64
ps_ind_11_bin     21694 non-null int64
ps_ind_12_bin     21694 non-null int64
ps_ind_13_bin     21694 non-null int64
ps_ind_14         21694 non-null int64
ps_ind_15         21694 non-null int64
ps_ind_16_bin     21694 non-null int64
ps_ind_17_bin     21694 non-null int64
ps_ind_18_bin     21694 non-null int64
ps_reg_01         21694 non-null float64
ps_reg_02         21694 non-null float64
ps_reg_03         21694 non-null float64
ps_car_01_cat     21694 non-null int64
ps_car_02_cat     21694 non-null int64
ps_car_03_cat     21694 non-null int64
ps_car_04_cat     21694 non-null int64
ps_car_05_cat     21694 non-null int64
ps_car_06_cat     21694 non-null int64
ps_car_07_cat     21694 non-null int64
ps_car_08_cat     21694 non-null int64
ps_car_09_cat     21694 non-null int64
ps_car_10_cat     21694 non-null int64
ps_car_11_cat     21694 non-null int64
ps_car_11         21694 non-null int64
ps_car_12         21694 non-null float64
ps_car_13         21694 non-null float64
ps_car_14         21694 non-null float64
ps_car_15         21694 non-null float64
ps_calc_01        21694 non-null float64
ps_calc_02        21694 non-null float64
ps_calc_03        21694 non-null float64
ps_calc_04        21694 non-null int64
ps_calc_05        21694 non-null int64
ps_calc_06        21694 non-null int64
ps_calc_07        21694 non-null int64
ps_calc_08        21694 non-null int64
ps_calc_09        21694 non-null int64
ps_calc_10        21694 non-null int64
ps_calc_11        21694 non-null int64
ps_calc_12        21694 non-null int64
ps_calc_13        21694 non-null int64
ps_calc_14        21694 non-null int64
ps_calc_15_bin    21694 non-null int64
ps_calc_16_bin    21694 non-null int64
ps_calc_17_bin    21694 non-null int64
ps_calc_18_bin    21694 non-null int64
ps_calc_19_bin    21694 non-null int64
ps_calc_20_bin    21694 non-null int64
dtypes: float64(10), int64(49)
memory usage: 9.9 MB

















In [31]:



    


# sample negatives
neg_data = neg_data.sample(n=21694)
neg_data.info()



    


















    






<class 'pandas.core.frame.DataFrame'>
Int64Index: 21694 entries, 564168 to 583460
Data columns (total 59 columns):
id                21694 non-null int64
target            21694 non-null int64
ps_ind_01         21694 non-null int64
ps_ind_02_cat     21694 non-null int64
ps_ind_03         21694 non-null int64
ps_ind_04_cat     21694 non-null int64
ps_ind_05_cat     21694 non-null int64
ps_ind_06_bin     21694 non-null int64
ps_ind_07_bin     21694 non-null int64
ps_ind_08_bin     21694 non-null int64
ps_ind_09_bin     21694 non-null int64
ps_ind_10_bin     21694 non-null int64
ps_ind_11_bin     21694 non-null int64
ps_ind_12_bin     21694 non-null int64
ps_ind_13_bin     21694 non-null int64
ps_ind_14         21694 non-null int64
ps_ind_15         21694 non-null int64
ps_ind_16_bin     21694 non-null int64
ps_ind_17_bin     21694 non-null int64
ps_ind_18_bin     21694 non-null int64
ps_reg_01         21694 non-null float64
ps_reg_02         21694 non-null float64
ps_reg_03         21694 non-null float64
ps_car_01_cat     21694 non-null int64
ps_car_02_cat     21694 non-null int64
ps_car_03_cat     21694 non-null int64
ps_car_04_cat     21694 non-null int64
ps_car_05_cat     21694 non-null int64
ps_car_06_cat     21694 non-null int64
ps_car_07_cat     21694 non-null int64
ps_car_08_cat     21694 non-null int64
ps_car_09_cat     21694 non-null int64
ps_car_10_cat     21694 non-null int64
ps_car_11_cat     21694 non-null int64
ps_car_11         21694 non-null int64
ps_car_12         21694 non-null float64
ps_car_13         21694 non-null float64
ps_car_14         21694 non-null float64
ps_car_15         21694 non-null float64
ps_calc_01        21694 non-null float64
ps_calc_02        21694 non-null float64
ps_calc_03        21694 non-null float64
ps_calc_04        21694 non-null int64
ps_calc_05        21694 non-null int64
ps_calc_06        21694 non-null int64
ps_calc_07        21694 non-null int64
ps_calc_08        21694 non-null int64
ps_calc_09        21694 non-null int64
ps_calc_10        21694 non-null int64
ps_calc_11        21694 non-null int64
ps_calc_12        21694 non-null int64
ps_calc_13        21694 non-null int64
ps_calc_14        21694 non-null int64
ps_calc_15_bin    21694 non-null int64
ps_calc_16_bin    21694 non-null int64
ps_calc_17_bin    21694 non-null int64
ps_calc_18_bin    21694 non-null int64
ps_calc_19_bin    21694 non-null int64
ps_calc_20_bin    21694 non-null int64
dtypes: float64(10), int64(49)
memory usage: 9.9 MB

















In [32]:



    


all_data = pd.concat([pos_data, neg_data])
all_data = all_data.sample(frac=1).reset_index(drop=True) # shuffle
all_data.info()



    


















    






<class 'pandas.core.frame.DataFrame'>
RangeIndex: 43388 entries, 0 to 43387
Data columns (total 59 columns):
id                43388 non-null int64
target            43388 non-null int64
ps_ind_01         43388 non-null int64
ps_ind_02_cat     43388 non-null int64
ps_ind_03         43388 non-null int64
ps_ind_04_cat     43388 non-null int64
ps_ind_05_cat     43388 non-null int64
ps_ind_06_bin     43388 non-null int64
ps_ind_07_bin     43388 non-null int64
ps_ind_08_bin     43388 non-null int64
ps_ind_09_bin     43388 non-null int64
ps_ind_10_bin     43388 non-null int64
ps_ind_11_bin     43388 non-null int64
ps_ind_12_bin     43388 non-null int64
ps_ind_13_bin     43388 non-null int64
ps_ind_14         43388 non-null int64
ps_ind_15         43388 non-null int64
ps_ind_16_bin     43388 non-null int64
ps_ind_17_bin     43388 non-null int64
ps_ind_18_bin     43388 non-null int64
ps_reg_01         43388 non-null float64
ps_reg_02         43388 non-null float64
ps_reg_03         43388 non-null float64
ps_car_01_cat     43388 non-null int64
ps_car_02_cat     43388 non-null int64
ps_car_03_cat     43388 non-null int64
ps_car_04_cat     43388 non-null int64
ps_car_05_cat     43388 non-null int64
ps_car_06_cat     43388 non-null int64
ps_car_07_cat     43388 non-null int64
ps_car_08_cat     43388 non-null int64
ps_car_09_cat     43388 non-null int64
ps_car_10_cat     43388 non-null int64
ps_car_11_cat     43388 non-null int64
ps_car_11         43388 non-null int64
ps_car_12         43388 non-null float64
ps_car_13         43388 non-null float64
ps_car_14         43388 non-null float64
ps_car_15         43388 non-null float64
ps_calc_01        43388 non-null float64
ps_calc_02        43388 non-null float64
ps_calc_03        43388 non-null float64
ps_calc_04        43388 non-null int64
ps_calc_05        43388 non-null int64
ps_calc_06        43388 non-null int64
ps_calc_07        43388 non-null int64
ps_calc_08        43388 non-null int64
ps_calc_09        43388 non-null int64
ps_calc_10        43388 non-null int64
ps_calc_11        43388 non-null int64
ps_calc_12        43388 non-null int64
ps_calc_13        43388 non-null int64
ps_calc_14        43388 non-null int64
ps_calc_15_bin    43388 non-null int64
ps_calc_16_bin    43388 non-null int64
ps_calc_17_bin    43388 non-null int64
ps_calc_18_bin    43388 non-null int64
ps_calc_19_bin    43388 non-null int64
ps_calc_20_bin    43388 non-null int64
dtypes: float64(10), int64(49)
memory usage: 19.5 MB

















In [33]:



    


# standardize data
df_x = all_data.iloc[:,2:]
df_y = all_data.iloc[:,:2]

scaler = StandardScaler().fit(df_x)
df_x = pd.DataFrame(scaler.transform(df_x))

all_data = pd.concat([df_y, df_x], axis=1)
all_data.head()



    


















    
Out[33]:










  
    

      
      id
      target
      0
      1
      2
      3
      4
      5
      6
      7
      ...
      47
      48
      49
      50
      51
      52
      53
      54
      55
      56
    

  
  
    

      0
      397909
      1
      0.006190
      -0.540064
      -0.176771
      1.149286
      -0.336169
      -0.735220
      1.537339
      -0.462444
      ...
      1.095924
      -0.361753
      -0.511807
      -0.564091
      -0.375711
      0.770888
      0.895072
      -0.632721
      1.376676
      -0.421670
    

    

      1
      993433
      0
      -0.491145
      -0.540064
      1.643982
      -0.864413
      0.324924
      1.360137
      -0.650475
      -0.462444
      ...
      -1.042133
      0.470427
      0.666527
      0.524612
      -0.375711
      0.770888
      -1.117228
      -0.632721
      1.376676
      -0.421670
    

    

      2
      804416
      1
      1.498193
      3.892554
      2.372283
      1.149286
      -0.336169
      -0.735220
      1.537339
      -0.462444
      ...
      1.095924
      -0.361753
      1.844861
      2.339118
      -0.375711
      0.770888
      0.895072
      -0.632721
      1.376676
      -0.421670
    

    

      3
      576857
      0
      -0.491145
      0.937475
      -0.176771
      1.149286
      -0.336169
      -0.735220
      -0.650475
      2.162426
      ...
      0.668313
      2.134787
      0.666527
      0.524612
      -0.375711
      -1.297206
      -1.117228
      -0.632721
      1.376676
      -0.421670
    

    

      4
      267011
      0
      -0.988479
      0.937475
      0.187380
      1.149286
      -0.336169
      -0.735220
      1.537339
      -0.462444
      ...
      0.668313
      2.134787
      0.077360
      0.161711
      -0.375711
      0.770888
      -1.117228
      -0.632721
      -0.726388
      2.371523
    

  



5 rows × 59 columns



















In [34]:



    


x_train = all_data.iloc[:40000,2:]
y_train = all_data.iloc[:40000,1]
x_test = all_data.iloc[40000:,2:]
y_test = all_data.iloc[40000:,1]
y_test.head()



    


















    
Out[34]:








40000    0
40001    1
40002    0
40003    1
40004    1
Name: target, dtype: int64

















In [39]:



    


mlp = MLPClassifier(hidden_layer_sizes=(300), 
                    activation='logistic')

mlp.fit(x_train, y_train)



    


















    
Out[39]:








MLPClassifier(activation='logistic', alpha=0.0001, batch_size='auto',
       beta_1=0.9, beta_2=0.999, early_stopping=False, epsilon=1e-08,
       hidden_layer_sizes=300, learning_rate='constant',
       learning_rate_init=0.001, max_iter=200, momentum=0.9,
       nesterovs_momentum=True, power_t=0.5, random_state=None,
       shuffle=True, solver='adam', tol=0.0001, validation_fraction=0.1,
       verbose=False, warm_start=False)

















In [40]:



    


mlp.score(x_test, y_test)



    


















    
Out[40]:








0.5832349468713105

















In [41]:



    


test_data = pd.read_csv('../kaggle_porto_seguro/test.csv')
test_data.head()



    


















    
Out[41]:










  
    

      
      id
      ps_ind_01
      ps_ind_02_cat
      ps_ind_03
      ps_ind_04_cat
      ps_ind_05_cat
      ps_ind_06_bin
      ps_ind_07_bin
      ps_ind_08_bin
      ps_ind_09_bin
      ...
      ps_calc_11
      ps_calc_12
      ps_calc_13
      ps_calc_14
      ps_calc_15_bin
      ps_calc_16_bin
      ps_calc_17_bin
      ps_calc_18_bin
      ps_calc_19_bin
      ps_calc_20_bin
    

  
  
    

      0
      0
      0
      1
      8
      1
      0
      0
      1
      0
      0
      ...
      1
      1
      1
      12
      0
      1
      1
      0
      0
      1
    

    

      1
      1
      4
      2
      5
      1
      0
      0
      0
      0
      1
      ...
      2
      0
      3
      10
      0
      0
      1
      1
      0
      1
    

    

      2
      2
      5
      1
      3
      0
      0
      0
      0
      0
      1
      ...
      4
      0
      2
      4
      0
      0
      0
      0
      0
      0
    

    

      3
      3
      0
      1
      6
      0
      0
      1
      0
      0
      0
      ...
      5
      1
      0
      5
      1
      0
      1
      0
      0
      0
    

    

      4
      4
      5
      1
      7
      0
      0
      0
      0
      0
      1
      ...
      4
      0
      0
      4
      0
      1
      1
      0
      0
      1
    

  



5 rows × 58 columns



















In [42]:



    


predictions = mlp.predict(test_data.iloc[:,1:])



    











In [43]:



    


submission = pd.DataFrame()
submission['id'] = test_data.iloc[:, 0]
submission['target'] = predictions
submission.head()



    


















    
Out[43]:










  
    

      
      id
      target
    

  
  
    

      0
      0
      0
    

    

      1
      1
      1
    

    

      2
      2
      0
    

    

      3
      3
      0
    

    

      4
      4
      1
    

  




















In [44]:



    


submission.to_csv('kaggle_submission_nn.csv', index=False)



    











In [ ]:

Content source: abevieiramota/data-science-cookbook

Similar notebooks:

notebook.community | gallery | about