如何在Python中打印班级预测？ （XGBoost）

Question

我已经完成了功能工程，并训练了我的XGBoost预测模型。我的准确度达到了81％，但我不知道如何建立混淆矩阵（例如：打印真实的客户流失率1-0与预测性流失率）。这可能吗？我想知道“好预测”的总和（正确的肯定和否定的真实）和“错误的预测”的总数量（错误的肯定和错误的否定）。我在下面分享我的代码。预先感谢。

    ### Training XGBoost Model
xgb_train = xgb.DMatrix(X_train, label=y_train)
clf = xgb.XGBClassifier(
            max_depth = 4,
            n_estimators = 600,
            learning_rate = 0.05, 
            nthread = 4,
            colsample_bytree = 0.5,
            min_child_weight = 5,
            seed = 0)

## Cross Validation
cv = xgb.cv(clf.get_xgb_params(), xgb_train,
        num_boost_round=600, 
        early_stopping_rounds=50, 
        nfold=5, 
        metrics=['auc'],
        seed=0)

# Fitting model with best parameters
clf.set_params(n_estimators=cv.shape[0])
clf.fit(X_train, y_train, eval_metric='auc')

# Calculate AUC in train and test.
auc_train = roc_auc_score(y_train, clf.predict_proba(X_train)[:,1])
y_proba_xgb = clf.predict_proba(X_test)[:,1]
auc_test = roc_auc_score(y_test, y_proba_xgb)
print(f"TRAIN: {auc_train}, TEST: {auc_test}")