如下图是该数据集正负样本的分布情况,异常情况相比正常情况极少。 面对非平衡数据集时,有两种解决方案:过采样和下采样。 下采样: 让数量多的样本减少到和数量少的样本数量一样多。 过采样:生成数量少的样本,以平衡数据。
下采样代码:
import numpy as np import pandas as pd import matplotlib.pyplot as plt from IPython import get_ipython get_ipython().run_line_magic('matplotlib', 'inline') data=pd.read_csv("creditcard.csv") data.head() #可以看到正负样本分布极不平衡,异常样本很少 count_classes=pd.value_counts(data["Class"],sort=True).sort_index() count_classes.plot(kind='bar') plt.title("Fraud class histogram") plt.xlabel("Class") plt.ylabel("Frequency") #sklearn库提供机器学习操作和预处理操作 #fit_transform将Amount特征变换为一列数据 from sklearn.preprocessing import StandardScaler data["normAmount"]=StandardScaler().fit_transform(data["Amount"].reshape(-1,1)) #删除Time和Amount两列 data=data.drop(["Time","Amount"],axis=1) #下采样 X = data.ix[:, data.columns != 'Class'] y = data.ix[:, data.columns == 'Class'] #统计异常样本数目和索引 number_records_fraud = len(data[data.Class == 1]) fraud_indices = np.array(data[data.Class == 1].index) normal_indices = data[data.Class == 0].index #random模块,随机选择和异常事件一样多的正常数据 random_normal_indices = np.random.choice(normal_indices, number_records_fraud, replace = False) random_normal_indices = np.array(random_normal_indices) #合并异常和正常的数据集 under_sample_indices = np.concatenate([fraud_indices,random_normal_indices]) 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'] 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))下采样导致数据量变少。虽然召回率能达到要求,但是将正常类错分为异常类的几率很高。如下图混淆矩阵所示,误分的有9895个,这就是下采样的劣势。
切分数据为两部分。80%训练,20%测试。平均切分训练集为三份1、2、3,分别用1、2来训练,3来验证;用1、3训练,2来验证;用2、3训练、1来验证。利用交叉验证找到合适的模型参数。 本案例采用逻辑斯特回归模型,在skleaarn库中有。
#Recall=TP/(TP+FN) #cross_val_score表示交叉验证的结果 #混淆矩阵 confusion_matrix from sklearn.linear_model import LogisticRegression from sklearn.cross_validation import KFold, cross_val_score from sklearn.metrics import confusion_matrix,recall_score,classification_report #5折交叉验证 def printing_Kfold_scores(x_train_data,y_train_data): fold=KFold(len(y_train_data),5,shuffle=False) #C就是正则化惩罚项中的λ 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 j=0 for c_param in c_param_range: print('-----------------') print('C parameter:',c_param) print('-----------------') print('') recall_accs=[] #enumerate枚举类型,代表从1开始,在fold中迭代 for iteration,indices in enumerate(fold,start=1): lr=LogisticRegression(C=c_param,penalty='l1')#选择了L1惩罚,L1、L2均可以 lr.fit(x_train_data.iloc[indices[0],:],y_train_data.iloc[indices[0],:].values.ravel()) y_pred_undersample=lr.predict(x_train_data.iloc[indices[1],:].values) 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.loc[results_table['Mean recall score'].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)模型有时候很容易就建立出来,但是如何确定模型的有效性和适用性就需要一个评估标准。 TP(true positives):将正类预测为正类数,true代表判断正确,positives为正类 FN(false negatives):将正类预测为负类数 FP(false positives):将负类预测为正类数 TN(true negatives):将负类预测为负类数 精确率: p=TPTP+FP 召回率: R=TPTP+FN 通常使用召回率作为评估模型标准。
L2正则化, 1/2w2 ,前面通常乘参数λ L1正则化, |w|
1.针对少数类中的每一个样本,找到其k个近邻(k值可选); 2.针对每一个样本,根据少数类需要扩大的情况,从近邻中随机挑选出需要的近邻; 例如,需要少数类增加200%,即原来有100个,希望扩到到300个,就从k个近邻中随机挑选出两个近邻。 3.针对一个样本a和它的近邻b: 3.1.计算两者在各个特征空间中值的差值; 3.2.并将这个差值乘以一个(0,1)的随机数后,与当前样本的特征值相加,作为新的合成样本。
首先需要人为地安装imblearn库,打开anaconda prompt输入指令:
pip install imblearnSMOTE生成样本代码如下:
##过采样SMOTE策略 import pandas as pd from imblearn.over_sampling import SMOTE from sklearn.ensemble import RandomForestClassifier from sklearn.metrics import confusion_matrix from sklearn.model_selection import train_test_split credit_cards=pd.read_csv('creditcard.csv') columns=credit_cards.columns # The labels are in the last column ('Class'). Simply remove it to obtain features columns features_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) #只对训练进行样本生成,另外的测试集不需要生成 oversampler=SMOTE(random_state=0) os_features,os_labels=oversampler.fit_sample(features_train,labels_train) os_features = pd.DataFrame(os_features) os_labels = pd.DataFrame(os_labels) best_c = printing_Kfold_scores(os_features,os_labels) lr = LogisticRegression(C = best_c, penalty = 'l1') lr.fit(os_features,os_labels.values.ravel()) y_pred = lr.predict(features_test.values) # Compute confusion matrix cnf_matrix = confusion_matrix(labels_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()过采样的混淆矩阵如图: 相比下采样效果更好,误分的数据较少。 一般来说数据越多,模型会越好,因此一般会采用过采样方法解决非平衡数据问题。
