ben_第四次作业

1、背景介绍：

C银行信用卡中心在对欺诈风险和反欺诈技术作了充分研究之后，融合内外部数据，建立以评分模型为支撑的欺诈识别和防范系统，以满足精细化管理的需要。本次作业根据提供的数据（“FRAUD_TRAIN_Samp.csv”，引用自陈春宝等出版的《SAS金融数据挖掘与建模》）建立信用卡申请反欺诈模型。

2、主要变量说明如下：

#无-组合算法为黑箱模型，无需知道变量含义

3、作业安排：

• 3.1 基础知识：

  1）比较逻辑回归、决策树、神经网络、组合算法的适用场景和优缺点。
  
  

• 3.2 案例解答步骤如下：

  1）使用决策树、神经网络、组合算法建立反欺诈模型，比较三个模型的表现。

  2)自学绘制PR曲线：横轴为精确度(Precise)，纵轴为召回率(Recall)。

In [34]:

import numpy as np

import pandas as pd

import matplotlib.pyplot as plt

import seaborn as sns

from scipy import stats

from  sklearn.tree import DecisionTreeClassifier

from sklearn.linear_model import LogisticRegression

from sklearn.model_selection import GridSearchCV

from sklearn import metrics

from sklearn.neural_network import MLPClassifier

​

%matplotlib inline

sample=pd.read_excel('samp.xlsx',encoding='utf-8')

In [35]:

sample.head()

Out[35]:

csr_idtargetF2F3F4F5F6F7F8F9...F237F238F239F240F241F242F243F244F245F246
085350100601100...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
173450000431118012...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
28437030090801...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
3616601600502310016...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
4937001400540004...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN

5 rows × 247 columns

In [36]:

sample.describe()

Out[36]:

csr_idtargetF2F3F4F5F6F7F8F9...F237F238F239F240F241F242F243F244F245F246
count972.000000972.000000972.000000972.000000972.000000972.000000972.000000972.000000972.000000972.000000...7.07.07.0000007.0000007.07.07.07.0000007.0000007.000000
mean5109.4650210.05041210.0668720.1460910.85802527.9907419.3014409.2407414.5257207.253086...0.00.071642.85714371642.8571430.00.00.083282.81428683711.38571412068.528571
std2894.0404750.21890520.5666551.33681713.72603830.70907518.49231420.66384518.31682412.743680...0.00.0145591.527422145591.5274220.00.00.0188748.601642188531.264044264389.597515
min20.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.0000000.000000...0.00.00.0000000.0000000.00.00.00.0000000.000000-400000.000000
25%2530.5000000.0000001.7500000.0000000.0000008.0000000.0000001.0000000.0000000.000000...0.00.03500.0000003500.0000000.00.00.00.0000001500.000000-13209.535000
50%5296.0000000.0000007.0000000.0000000.00000020.0000003.0000004.0000000.0000003.000000...0.00.025000.00000025000.0000000.00.00.012780.93000012780.930000-7000.000000
75%7628.2500000.00000012.0000000.0000000.00000039.00000011.00000012.0000003.00000010.000000...0.00.034750.00000034750.0000000.00.00.030150.00000030150.0000004000.000000
max9996.0000001.000000483.00000030.000000427.000000327.000000267.000000538.000000457.000000262.000000...0.00.0400000.000000400000.0000000.00.00.0509898.770000509898.770000509898.770000

8 rows × 243 columns

In [37]:

sns.distplot(sample.target,kde=True,fit=stats.norm)

Out[37]:

<matplotlib.axes._subplots.AxesSubplot at 0x23d3bdfebe0>

In [38]:

sample_1=sample.loc[sample.iloc[:,1]==1,:]

sample_0=sample.loc[sample.iloc[:,1]==0,:]

In [39]:

sample_1.describe()

Out[39]:

csr_idtargetF2F3F4F5F6F7F8F9...F237F238F239F240F241F242F243F244F245F246
count49.00000049.049.00000049.00000049.00000049.00000049.00000049.00000049.00000049.000000...1.01.01.01.01.01.01.01.01.01.0
mean251.8571431.08.0612240.2040820.71428631.30612210.1632659.3469392.6530617.714286...0.00.0400000.0400000.00.00.00.00.00.0-400000.0
std129.9474250.07.5482010.7065051.48604623.44426418.25210810.5347665.3211499.438397...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
min20.0000001.00.0000000.0000000.0000000.0000000.0000000.0000000.0000000.000000...0.00.0400000.0400000.00.00.00.00.00.0-400000.0
259.0000001.02.0000000.0000000.00000014.0000000.0000001.0000000.0000000.000000...0.00.0400000.0400000.00.00.00.00.00.0-400000.0
50%233.0000001.07.0000000.0000000.00000027.0000004.0000007.0000000.0000004.000000...0.00.0400000.0400000.00.00.00.00.00.0-400000.0
75%372.0000001.012.0000000.0000000.00000043.00000012.00000013.0000002.00000013.000000...0.00.0400000.0400000.00.00.00.00.00.0-400000.0
max485.0000001.038.0000003.0000007.000000109.000000113.00000047.00000027.00000031.000000...0.00.0400000.0400000.00.00.00.00.00.0-400000.0

8 rows × 243 columns

In [40]:

##数据预处理F81 F74 F75

from pylab import mpl

mpl.rcParams['font.sans-serif']=['SimHei']#指定默认字体

mpl.rcParams['axes.unicode_minus']=False

sample.F81=sample.F81.fillna('缺失')

sample.F81.value_counts().plot(kind='bar')

dict81={'无等级':1,'缺失':2,'平衡':3,'稳健':4,'成长':5,'进取':6,'保守':7}

sample.F81=sample.F81.map(dict81)

sample.F74.value_counts().plot(kind='bar')

dict74={'lvyou':1,'feiji':2,'tielu':3,'xchen':4,'JIUDIAN':5,'quna':6,'JYY':7}

sample.F74=sample.F81.map(dict74)

In [41]:

sample.F75=sample.F75.fillna('O')

sample.F75.unique()

​

Out[41]:

array(['M', 'F', 'O'], dtype=object)

In [42]:

sample.F75.value_counts().plot(kind='bar')

dict75={'M':0,'F':1,'O':2}

sample.F75=sample.F75.map(dict75)

In [43]:

#针对空值转化为0,针对非数字型记录其索引

cloindex=[]

for i in range(245):

    sample.iloc[:,i+2]=sample.iloc[:,i+2].fillna(0)

sample=sample.drop('F80',1)

In [44]:

sample.head()

Out[44]:

csr_idtargetF2F3F4F5F6F7F8F9...F237F238F239F240F241F242F243F244F245F246
085350100601100...0.00.00.00.00.00.00.00.00.00.0
173450000431118012...0.00.00.00.00.00.00.00.00.00.0
28437030090801...0.00.00.00.00.00.00.00.00.00.0
3616601600502310016...0.00.00.00.00.00.00.00.00.00.0
4937001400540004...0.00.00.00.00.00.00.00.00.00.0

5 rows × 246 columns

In [45]:

sample.target.value_counts().plot(kind='bar')

Out[45]:

<matplotlib.axes._subplots.AxesSubplot at 0x23d3c00d748>

使用决策树构建模型

In [46]:

from sklearn import cross_validation

​

X_data=sample.iloc[:,2:]

y_data=sample.iloc[:,1]

E:\Anaconda3\lib\site-packages\sklearn\cross_validation.py:41: DeprecationWarning: This module was deprecated in version 0.18 in favor of the model_selection module into which all the refactored classes and functions are moved. Also note that the interface of the new CV iterators are different from that of this module. This module will be removed in 0.20.

  "This module will be removed in 0.20.", DeprecationWarning)

In [47]:

X_train,X_test,y_train,y_test=cross_validation.train_test_split(

    X_data,y_data,test_size=0.35,random_state=1

)

In [48]:

regr=DecisionTreeClassifier()

regr.fit(X_train,y_train)

print('Training score:%f'%(regr.score(X_train,y_train)))

print('Testing score:%f'%(regr.score(X_test,y_test)))

Training score:1.000000

Testing score:0.891496

In [49]:

result_train=regr.predict(X_train)

result_train_proba=regr.predict_proba(X_train)[:,1]

​

result_test=regr.predict(X_test)

result_test_proba=regr.predict_proba(X_test)[:,1]

​

pd.DataFrame({

    'y_test':y_test,

    'result_test':result_test,

    'result_test_proba':result_test_proba

}).T

Out[49]:

51756020569824194431132934133...921149527695435513257355458
y_test0.00.00.00.00.00.01.01.00.00.0...0.00.00.01.01.00.00.00.00.00.0
result_test0.00.00.00.00.01.00.00.00.00.0...0.00.00.00.00.00.00.00.00.00.0
result_test_proba0.00.00.00.00.01.00.00.00.00.0...0.00.00.00.00.00.00.00.00.00.0

3 rows × 341 columns

In [61]:

import sklearn.metrics as metrics

print(metrics.confusion_matrix(y_test,result_test,labels=[0,1]))

print(metrics.classification_report(y_test,result_test))

#print(pd.DataFrame(list(zip(sample.columns,regr.feature_importances_))))

[[301  22]

 [ 15   3]]

             precision    recall  f1-score   support



          0       0.95      0.93      0.94       323

          1       0.12      0.17      0.14        18



avg / total       0.91      0.89      0.90       341

In [51]:

import seaborn as sns

red,blue=sns.color_palette('Set1',2)

sns.distplot(result_test_proba[y_test==1],kde=False,bins=15,color=red)

sns.distplot(result_test_proba[y_test==0],kde=False,bins=15,color=blue)

plt.show()

In [59]:

#fpr_test,tpr_test,th_test=metrics.roc_curve(result_test,result_test_proba)

#fpr_train,tpr_train,th_train=metrics.roc_curve(result_train,result_train_proba)

fpr_test,tpr_test,th_test=metrics.roc_curve(y_test,result_test)

fpr_train,tpr_train,th_train=metrics.roc_curve(y_train,result_train)

plt.figure(figsize=[6,6])

plt.plot(fpr_test,tpr_test,color='blue')

plt.plot(fpr_train,tpr_train,color='red')

plt.title('ROC  curve')

plt.show()

In [62]:

p_test,r_test,_=metrics.precision_recall_curve(y_test,result_test)

p_train,r_train,_=metrics.precision_recall_curve(y_train,result_train)

plt.figure(figsize=[6,6])

plt.plot(r_test,p_test,color='blue')

plt.plot(r_train,p_train,color='red')

plt.title('P-R  曲线')

plt.show()

In [56]:

from sklearn.model_selection import GridSearchCV

from sklearn import metrics

In [58]:

param_grid={

    'max_depth':[2,3,4,5,6,7,8],

    'min_samples_split':[4,8,12,16,20,24,28]

}

clf=DecisionTreeClassifier(criterion='entropy')

clfcv=GridSearchCV(estimator=clf,param_grid=param_grid,

                  scoring='roc_auc',cv=4)

clfcv.fit(X_train,y_train)

Out[58]:

GridSearchCV(cv=4, error_score='raise',

       estimator=DecisionTreeClassifier(class_weight=None, criterion='entropy', max_depth=None,

            max_features=None, max_leaf_nodes=None,

            min_impurity_decrease=0.0, min_impurity_split=None,

            min_samples_leaf=1, min_samples_split=2,

            min_weight_fraction_leaf=0.0, presort=False, random_state=None,

            splitter='best'),

       fit_params=None, iid=True, n_jobs=1,

       param_grid={'max_depth': [2, 3, 4, 5, 6, 7, 8], 'min_samples_split': [4, 8, 12, 16, 20, 24, 28]},

       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',

       scoring='roc_auc', verbose=0)

In [ ]:

​

In [ ]:

​

In [158]:

mlp=MLPClassifier(hidden_layer_sizes=(10,),

                 activation='logistic',

                 alpha=0.1,

                 max_iter=100000)

mlp.fit(X_train,y_train)

mlp

Out[158]:

MLPClassifier(activation='logistic', alpha=0.1, batch_size='auto', beta_1=0.9,

       beta_2=0.999, early_stopping=False, epsilon=1e-08,

       hidden_layer_sizes=(10,), learning_rate='constant',

       learning_rate_init=0.001, max_iter=100000, 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 [144]:

train_predict=mlp.predict()

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

NameError                                 Traceback (most recent call last)

<ipython-input-144-5802580b8a93> in <module>()

----> 1 type(sample.iloc[5,2])==flaot



NameError: name 'flaot' is not defined

In [164]:

from sklearn.model_selection import GridSearchCV

from sklearn import metrics

param_grid={

    'max_iter':[1000,2000,3000,4000,5000],

    'hidden_layer_sizes':[(10,),(15,),(20,),(5,5)],

    'activation':['identity','logistic','tanh','relu'],

    'alpha':[0.001,0.01,0.1,0.2,0.4,1,10]

}

mlp=MLPClassifier()

gcv=GridSearchCV(estimator=mlp,param_grid=param_grid,

                scoring='roc_auc',cv=4,n_jobs=1)

gcv.fit(X_train,y_train)

Out[164]:

GridSearchCV(cv=4, error_score='raise',

       estimator=MLPClassifier(activation='relu', alpha=0.0001, batch_size='auto', beta_1=0.9,

       beta_2=0.999, early_stopping=False, epsilon=1e-08,

       hidden_layer_sizes=(100,), 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),

       fit_params=None, iid=True, n_jobs=1,

       param_grid={'max_iter': [1000, 2000, 3000, 4000, 5000], 'hidden_layer_sizes': [(10,), (15,), (20,), (5, 5)], 'activation': ['identity', 'logistic', 'tanh', 'relu'], 'alpha': [0.001, 0.01, 0.1, 0.2, 0.4, 1, 10]},

       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',

       scoring='roc_auc', verbose=0)