如何在交叉验证和GridSearchCV中实现SMOTE

我对 Python 比较陌生。您能帮助我将 SMOTE 的实施改进为适当的管道吗？我想要的是在每次 k 次迭代的训练集上应用过采样和欠采样，以便模型在平衡的数据集上进行训练，并在不平衡的遗漏部分上进行评估。问题是，当我这样做时，我无法使用熟悉的sklearn用于评估和网格搜索的界面。

是否可以做类似的东西model_selection.RandomizedSearchCV。我对此的看法：

df = pd.read_csv("Imbalanced_data.csv") #Load the data set
X = df.iloc[:,0:64]
X = X.values
y = df.iloc[:,64]
y = y.values
n_splits = 2
n_measures = 2 #Recall and AUC
kf = StratifiedKFold(n_splits=n_splits) #Stratified because we need balanced samples
kf.get_n_splits(X)
clf_rf = RandomForestClassifier(n_estimators=25, random_state=1)
s =(n_splits,n_measures)
scores = np.zeros(s)
for train_index, test_index in kf.split(X,y):
   print("TRAIN:", train_index, "TEST:", test_index)
   X_train, X_test = X[train_index], X[test_index]
   y_train, y_test = y[train_index], y[test_index]
   sm = SMOTE(ratio = 'auto',k_neighbors = 5, n_jobs = -1)
   smote_enn = SMOTEENN(smote = sm)
   x_train_res, y_train_res = smote_enn.fit_sample(X_train, y_train)
   clf_rf.fit(x_train_res, y_train_res)
   y_pred = clf_rf.predict(X_test,y_test)
   scores[test_index,1] = recall_score(y_test, y_pred)
   scores[test_index,2] = auc(y_test, y_pred)

您需要查看管道对象。不平衡学习有一个Pipeline https://imbalanced-learn.org/stable/generated/imblearn.pipeline.Pipeline.html它扩展了 scikit-learn Pipeline，除了 scikit-learn 的 fit_predict()、fit_transform() 和 Predict() 方法之外，还适应 fit_sample() 和 sample() 方法。

在这里看看这个例子：

• https://imbalanced-learn.org/stable/auto_examples/pipeline/plot_pipeline_classification.html https://imbalanced-learn.org/stable/auto_examples/pipeline/plot_pipeline_classification.html

对于您的代码，您需要这样做：

from imblearn.pipeline import make_pipeline, Pipeline

smote_enn = SMOTEENN(smote = sm)
clf_rf = RandomForestClassifier(n_estimators=25, random_state=1)

pipeline = make_pipeline(smote_enn, clf_rf)
    OR
pipeline = Pipeline([('smote_enn', smote_enn),
                     ('clf_rf', clf_rf)])

然后你可以通过这个pipeline对象 GridSearchCV、RandomizedSearchCV 或 scikit-learn 中的其他交叉验证工具作为常规对象。

kf = StratifiedKFold(n_splits=n_splits)
random_search = RandomizedSearchCV(pipeline, param_distributions=param_dist,
                                   n_iter=1000, 
                                   cv = kf)