信用卡欺诈行为逻辑回归数据分析-大数据ML样本集案例实战-白红宇

1 信用卡欺诈行为案例集预处理

import pandas as pdimport matplotlib.pyplot as pltimport numpy as np%matplotlib inlinedata = pd.read_csv("creditcard.csv")data.head()复制代码

from sklearn.preprocessing import StandardScalerdata['normAmount'] = StandardScaler().fit_transform(data['Amount'].values.reshape(-1, 1))data = data.drop(['Time','Amount'],axis=1)data.head()复制代码

2 K折交叉验证(验证通过)

def printing_Kfold_scores(x_train_data, y_train_data):        fold = KFold(5,shuffle=False)     # Different C parameters    # 0.01 倒数其实是100     # 0.1其实是10    c_param_range = [0.01,0.1,1,10,100]    results_table = pd.DataFrame(index = range(len(c_param_range),2), columns = ['C_parameter','Mean recall score'])    results_table['C_parameter'] = c_param_range    # the k-fold will give 2 lists: train_indices = indices[0], test_indices = indices[1]    j = 0    for c_param in c_param_range:        print('-------------------------------------------')        print('C parameter: ', c_param)        print('-------------------------------------------')        print('')        recall_accs = []        for iteration, indices in enumerate(fold.split(x_train_data)):            # Call the logistic regression model with a certain C parameter            lr = LogisticRegression(C = c_param, penalty = 'l1')            # Use the training data to fit the model. In this case, we use the portion of the fold to train the model            # with indices[0]. We then predict on the portion assigned as the 'test cross validation' with indices[1]            lr.fit(x_train_data.iloc[indices[0],:],y_train_data.iloc[indices[0],:].values.ravel())            # Predict values using the test indices in the training data            y_pred_undersample = lr.predict(x_train_data.iloc[indices[1],:].values)            # Calculate the recall score and append it to a list for recall scores representing the current c_parameter            recall_acc = recall_score(y_train_data.iloc[indices[1],:].values,y_pred_undersample)            recall_accs.append(recall_acc)            print('Iteration ', iteration,': recall score = ', recall_acc)        # The mean value of those recall scores is the metric we want to save and get hold of.        results_table.ix[j,'Mean recall score'] = np.mean(recall_accs)        j += 1        print('')        print('Mean recall score ', np.mean(recall_accs))        print('')        best_c = results_table.iloc[results_table['Mean recall score'].astype('float64').idxmax()]['C_parameter']        # Finally, we can check which C parameter is the best amongst the chosen.    print('*********************************************************************************')    print('Best model to choose from cross validation is with C parameter = ', best_c)    print('*********************************************************************************')        return best_c      best_c = printing_Kfold_scores(X_train_undersample,y_train_undersample)    -------------------------------------------    C parameter:  0.01    -------------------------------------------        Iteration  0 : recall score =  0.8082191780821918    Iteration  1 : recall score =  0.8356164383561644    Iteration  2 : recall score =  0.8983050847457628    Iteration  3 : recall score =  0.8918918918918919    Iteration  4 : recall score =  0.8939393939393939        Mean recall score  0.8655943974030809        -------------------------------------------    C parameter:  0.1    -------------------------------------------        Iteration  0 : recall score =  0.863013698630137    Iteration  1 : recall score =  0.8767123287671232    Iteration  2 : recall score =  0.9830508474576272    Iteration  3 : recall score =  0.918918918918919    Iteration  4 : recall score =  0.9090909090909091        Mean recall score  0.9101573405729431        -------------------------------------------    C parameter:  1    -------------------------------------------        Iteration  0 : recall score =  0.8767123287671232    Iteration  1 : recall score =  0.9041095890410958    Iteration  2 : recall score =  0.9830508474576272    Iteration  3 : recall score =  0.9459459459459459    Iteration  4 : recall score =  0.9242424242424242        Mean recall score  0.9268122270908433        -------------------------------------------    C parameter:  10    -------------------------------------------        Iteration  0 : recall score =  0.8904109589041096    Iteration  1 : recall score =  0.9041095890410958    C:\ProgramData\Anaconda3\lib\site-packages\ipykernel_launcher.py:39: DeprecationWarning:     .ix is deprecated. Please use    .loc for label based indexing or    .iloc for positional indexing        See the documentation here:    http://pandas.pydata.org/pandas-docs/stable/indexing.html#ix-indexer-is-deprecated    Iteration  2 : recall score =  0.9830508474576272    Iteration  3 : recall score =  0.9324324324324325    Iteration  4 : recall score =  0.9393939393939394        Mean recall score  0.9298795534458411        -------------------------------------------    C parameter:  100    -------------------------------------------        Iteration  0 : recall score =  0.9041095890410958    Iteration  1 : recall score =  0.9041095890410958    Iteration  2 : recall score =  0.9830508474576272    Iteration  3 : recall score =  0.9459459459459459    Iteration  4 : recall score =  0.9545454545454546        Mean recall score  0.9383522852062439        *********************************************************************************    Best model to choose from cross validation is with C parameter =  100.0    ********************************************************************************复制代码

3 不均衡问题处理策略（OverSample与UnderSample）

# 找出非class列    X = data.ix[:, data.columns != 'Class']    # 找出class列    y = data.ix[:, data.columns == 'Class']        # 找出欺诈的个数和索引492    number_records_fraud = len(data[data.Class == 1])    fraud_indices = np.array(data[data.Class == 1].index)        # Picking the indices of the normal classes（找出正常的索引）    normal_indices = data[data.Class == 0].index        # Out of the indices we picked, randomly select "x" number (number_records_fraud)（从正常的行为中选择接近欺诈的样本索引）492    random_normal_indices = np.random.choice(normal_indices, number_records_fraud, replace = False)    random_normal_indices = np.array(random_normal_indices)        # Appending the 2 indices(索引组合) 892    under_sample_indices = np.concatenate([fraud_indices,random_normal_indices])        # iloc通过行号获取行数据    under_sample_data = data.iloc[under_sample_indices,:]        X_undersample = under_sample_data.ix[:, under_sample_data.columns != 'Class']    y_undersample = under_sample_data.ix[:, under_sample_data.columns == 'Class']        # Showing ratio    print("Percentage of normal transactions: ", len(under_sample_data[under_sample_data.Class == 0])/len(under_sample_data))    print("Percentage of fraud transactions: ", len(under_sample_data[under_sample_data.Class == 1])/len(under_sample_data))    print("Total number of transactions in resampled data: ", len(under_sample_data))    Percentage of normal transactions:  0.5    Percentage of fraud transactions:  0.5    Total number of transactions in resampled data:  984复制代码

4 训练集与测试集划分

from sklearn.cross_validation import train_test_split        X特征输入,y表示label,test_size划分的测试集比例，没有设置random_state，每次取得的    结果就不一样，它的随机数种子与当前系统时间有关。其实就是该组随机数的编号，在需要重    复试验的时候，保证得到一组一样的随机数。比如你每次都填1，其他参数一样的情况下你得到    随机数组是一样的。但填0或不填，每次都不一样。随机数的产生取决于种子，随机数和种子之    间的关系遵从以下两个规则：种子不同，产生不同的随机数；种子相同，即使实例不同也产生    相同的随机数。        全部样本拆分    X_train, X_test, y_train, y_test = train_test_split(X,y,test_size = 0.3, random_state = 0)        print("Number transactions train dataset: ", len(X_train))    print("Number transactions test dataset: ", len(X_test))    print("Total number of transactions: ", len(X_train)+len(X_test))        Number transactions train dataset:  199364    Number transactions test dataset:  85443    Total number of transactions:  284807       # Undersampled dataset     X_train_undersample, X_test_undersample, y_train_undersample, y_test_undersample = train_test_split(X_undersample , y_undersample, test_size = 0.3, random_state = 0)        print("")    print("Number transactions train dataset: ", len(X_train_undersample))    print("Number transactions test dataset: ", len(X_test_undersample))    print("Total number of transactions: ", len(X_train_undersample)+len(X_test_undersample))        Number transactions train dataset:  688    Number transactions test dataset:  296    Total number of transactions:  984复制代码

5基于低采样数据集X_test_undersample模型训练与测试（均衡数据）

#Recall = TP/(TP+FN)from sklearn.linear_model import LogisticRegressionfrom sklearn.cross_validation import KFold, cross_val_scorefrom sklearn.metrics import confusion_matrix,recall_score,classification_report  函数调用 best_c = printing_Kfold_scores(X_train_undersample,y_train_undersample) -------------------------------------------C parameter:  0.01-------------------------------------------Iteration  1 : recall score =  0.958904109589Iteration  2 : recall score =  0.917808219178Iteration  3 : recall score =  1.0Iteration  4 : recall score =  0.972972972973Iteration  5 : recall score =  0.954545454545Mean recall score  0.960846151257-------------------------------------------C parameter:  0.1-------------------------------------------Iteration  1 : recall score =  0.835616438356Iteration  2 : recall score =  0.86301369863Iteration  3 : recall score =  0.915254237288Iteration  4 : recall score =  0.932432432432Iteration  5 : recall score =  0.878787878788Mean recall score  0.885020937099-------------------------------------------C parameter:  1-------------------------------------------Iteration  1 : recall score =  0.835616438356Iteration  2 : recall score =  0.86301369863Iteration  3 : recall score =  0.966101694915Iteration  4 : recall score =  0.945945945946Iteration  5 : recall score =  0.893939393939Mean recall score  0.900923434357-------------------------------------------C parameter:  10-------------------------------------------Iteration  1 : recall score =  0.849315068493Iteration  2 : recall score =  0.86301369863Iteration  3 : recall score =  0.966101694915Iteration  4 : recall score =  0.959459459459Iteration  5 : recall score =  0.893939393939Mean recall score  0.906365863087-------------------------------------------C parameter:  100-------------------------------------------Iteration  1 : recall score =  0.86301369863Iteration  2 : recall score =  0.86301369863Iteration  3 : recall score =  0.966101694915Iteration  4 : recall score =  0.959459459459Iteration  5 : recall score =  0.893939393939Mean recall score  0.909105589115*********************************************************************************Best model to choose from cross validation is with C parameter =  0.01*********************************************************************************复制代码

5 混合矩阵

def plot_confusion_matrix(cm, classes,                          title='Confusion matrix',                          cmap=plt.cm.Blues):    """    This function prints and plots the confusion matrix.    """    plt.imshow(cm, interpolation='nearest', cmap=cmap)    plt.title(title)    plt.colorbar()    tick_marks = np.arange(len(classes))    plt.xticks(tick_marks, classes, rotation=0)    plt.yticks(tick_marks, classes)    thresh = cm.max() / 2.    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):        plt.text(j, i, cm[i, j],                 horizontalalignment="center",                 color="white" if cm[i, j] > thresh else "black")    plt.tight_layout()    plt.ylabel('True label')    plt.xlabel('Predicted label')复制代码

6 混合矩阵作用于低采样数据集X_test_undersample的展示

import itertools    lr = LogisticRegression(C = best_c, penalty = 'l1')    lr.fit(X_train_undersample,y_train_undersample.values.ravel())    y_pred_undersample = lr.predict(X_test_undersample.values)        # Compute confusion matrix    cnf_matrix = confusion_matrix(y_test_undersample,y_pred_undersample)    np.set_printoptions(precision=2)        print("Recall metric in the testing dataset: ", cnf_matrix[1,1]/(cnf_matrix[1,0]+cnf_matrix[1,1]))        # Plot non-normalized confusion matrix    class_names = [0,1]    plt.figure()    plot_confusion_matrix(cnf_matrix                          , classes=class_names                          , title='Confusion matrix')    plt.show()复制代码

7 混合矩阵作用于全数据集X_test.values的展示

lr = LogisticRegression(C = best_c, penalty = 'l1')    lr.fit(X_train_undersample,y_train_undersample.values.ravel())    y_pred = lr.predict(X_test.values)        # Compute confusion matrix    cnf_matrix = confusion_matrix(y_test,y_pred)    np.set_printoptions(precision=2)        print("Recall metric in the testing dataset: ", cnf_matrix[1,1]/(cnf_matrix[1,0]+cnf_matrix[1,1]))        # Plot non-normalized confusion matrix    class_names = [0,1]    plt.figure()    plot_confusion_matrix(cnf_matrix                          , classes=class_names                          , title='Confusion matrix')    plt.show()复制代码

8 基于全数据集进行k折交叉验证（不均衡数据）

8.1 全数据集进行k折交叉验证

best_c = printing_Kfold_scores(X_train,y_train)-------------------------------------------C parameter:  0.01-------------------------------------------Iteration  1 : recall score =  0.492537313433Iteration  2 : recall score =  0.602739726027Iteration  3 : recall score =  0.683333333333Iteration  4 : recall score =  0.569230769231Iteration  5 : recall score =  0.45Mean recall score  0.559568228405-------------------------------------------C parameter:  0.1-------------------------------------------Iteration  1 : recall score =  0.567164179104Iteration  2 : recall score =  0.616438356164Iteration  3 : recall score =  0.683333333333Iteration  4 : recall score =  0.584615384615Iteration  5 : recall score =  0.525Mean recall score  0.595310250644-------------------------------------------C parameter:  1-------------------------------------------Iteration  1 : recall score =  0.55223880597Iteration  2 : recall score =  0.616438356164Iteration  3 : recall score =  0.716666666667Iteration  4 : recall score =  0.615384615385Iteration  5 : recall score =  0.5625Mean recall score  0.612645688837-------------------------------------------C parameter:  10-------------------------------------------Iteration  1 : recall score =  0.55223880597Iteration  2 : recall score =  0.616438356164Iteration  3 : recall score =  0.733333333333Iteration  4 : recall score =  0.615384615385Iteration  5 : recall score =  0.575Mean recall score  0.61847902217-------------------------------------------C parameter:  100-------------------------------------------Iteration  1 : recall score =  0.55223880597Iteration  2 : recall score =  0.616438356164Iteration  3 : recall score =  0.733333333333Iteration  4 : recall score =  0.615384615385Iteration  5 : recall score =  0.575Mean recall score  0.61847902217*********************************************************************************Best model to choose from cross validation is with C parameter =  10.0*********************************************************************************复制代码

8.2 全数据集混合矩阵

# 不均衡样本偏向于多的样本，误伤率低lr = LogisticRegression(C = best_c, penalty = 'l1')lr.fit(X_train,y_train.values.ravel())y_pred_undersample = lr.predict(X_test.values)# Compute confusion matrixcnf_matrix = confusion_matrix(y_test,y_pred_undersample)np.set_printoptions(precision=2)print("Recall metric in the testing dataset: ", cnf_matrix[1,1]/(cnf_matrix[1,0]+cnf_matrix[1,1]))# Plot non-normalized confusion matrixclass_names = [0,1]plt.figure()plot_confusion_matrix(cnf_matrix                      , classes=class_names                      , title='Confusion matrix')plt.show()复制代码

9 逻辑回归基于阈值进行判断（概率）

lr = LogisticRegression(C = 0.01, penalty = 'l1')lr.fit(X_train_undersample,y_train_undersample.values.ravel())y_pred_undersample_proba = lr.predict_proba(X_test_undersample.values)thresholds = [0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9]plt.figure(figsize=(10,10))j = 1for i in thresholds:    y_test_predictions_high_recall = y_pred_undersample_proba[:,1] > i        plt.subplot(3,3,j)    j += 1        # Compute confusion matrix    cnf_matrix = confusion_matrix(y_test_undersample,y_test_predictions_high_recall)    np.set_printoptions(precision=2)    print("Recall metric in the testing dataset: ", cnf_matrix[1,1]/(cnf_matrix[1,0]+cnf_matrix[1,1]))    # Plot non-normalized confusion matrix    class_names = [0,1]    plot_confusion_matrix(cnf_matrix                          , classes=class_names                          , title='Threshold >= %s'%i)  Recall metric in the testing dataset:  1.0Recall metric in the testing dataset:  1.0Recall metric in the testing dataset:  1.0Recall metric in the testing dataset:  0.986394557823Recall metric in the testing dataset:  0.931972789116Recall metric in the testing dataset:  0.884353741497Recall metric in the testing dataset:  0.836734693878Recall metric in the testing dataset:  0.748299319728Recall metric in the testing dataset:  0.571428571429 复制代码

10 基于SMOTE 进行数据预处理

import pandas as pdfrom imblearn.over_sampling import SMOTEfrom sklearn.ensemble import RandomForestClassifierfrom sklearn.metrics import confusion_matrixfrom sklearn.model_selection import train_test_splitcredit_cards=pd.read_csv('creditcard.csv')columns=credit_cards.columns# The labels are in the last column ('Class'). Simply remove it to obtain features columnsfeatures_columns=columns.delete(len(columns)-1)features=credit_cards[features_columns]labels=credit_cards['Class']features_train, features_test, labels_train, labels_test = train_test_split(features,                                                                         labels,                                                                         test_size=0.2,                                                                         random_state=0)oversampler=SMOTE(random_state=0)os_features,os_labels=oversampler.fit_sample(features_train,labels_train)len(os_labels[os_labels==1])227454os_features = pd.DataFrame(os_features)os_labels = pd.DataFrame(os_labels)best_c = printing_Kfold_scores(os_features,os_labels)-------------------------------------------C parameter:  0.01-------------------------------------------Iteration  1 : recall score =  0.890322580645Iteration  2 : recall score =  0.894736842105Iteration  3 : recall score =  0.968861347792Iteration  4 : recall score =  0.957595541926Iteration  5 : recall score =  0.958430881173Mean recall score  0.933989438728-------------------------------------------C parameter:  0.1-------------------------------------------Iteration  1 : recall score =  0.890322580645Iteration  2 : recall score =  0.894736842105Iteration  3 : recall score =  0.970410534469Iteration  4 : recall score =  0.959980655302Iteration  5 : recall score =  0.960178498807Mean recall score  0.935125822266-------------------------------------------C parameter:  1-------------------------------------------Iteration  1 : recall score =  0.890322580645Iteration  2 : recall score =  0.894736842105Iteration  3 : recall score =  0.970454796946Iteration  4 : recall score =  0.96014552489Iteration  5 : recall score =  0.960596168431Mean recall score  0.935251182603-------------------------------------------C parameter:  10-------------------------------------------Iteration  1 : recall score =  0.890322580645Iteration  2 : recall score =  0.894736842105Iteration  3 : recall score =  0.97065397809Iteration  4 : recall score =  0.960343368396Iteration  5 : recall score =  0.960530220596Mean recall score  0.935317397966-------------------------------------------C parameter:  100-------------------------------------------Iteration  1 : recall score =  0.890322580645Iteration  2 : recall score =  0.894736842105Iteration  3 : recall score =  0.970543321899Iteration  4 : recall score =  0.960211472725Iteration  5 : recall score =  0.960903924995Mean recall score  0.935343628474*********************************************************************************Best model to choose from cross validation is with C parameter =  100.0*********************************************************************************lr = LogisticRegression(C = best_c, penalty = 'l1')lr.fit(os_features,os_labels.values.ravel())y_pred = lr.predict(features_test.values)# Compute confusion matrixcnf_matrix = confusion_matrix( ,y_pred)np.set_printoptions(precision=2)print("Recall metric in the testing dataset: ", cnf_matrix[1,1]/(cnf_matrix[1,0]+cnf_matrix[1,1]))# Plot non-normalized confusion matrixclass_names = [0,1]plt.figure()plot_confusion_matrix(cnf_matrix                      , classes=class_names                      , title='Confusion matrix')plt.show()复制代码

11 总结

OverSample与UnderSample对比发现，基于SMOTE，数据的准确率和召回率得到了很大程度的提高。

秦凯新于深圳 201812081811