gbdt模型训练遇到报错

Terence888

代码如下

1. #Gradient Boost Decision Tree GBDT
   
2. from sklearn.ensemble import GradientBoostingClassifier
   
3. from sklearn.model_selection import GridSearchCV
   
4. from sklearn.metrics import accuracy_score
   
5. 
   
6. gbc = GradientBoostingClassifier()
   
7. param_test = {'n_estimators': [50, 100],
   
8.               'min_samples_split': [3, 4, 5, 6, 7],
   
9.               'max_depth': [3, 4, 5, 6]}
   
10. gbc_grid = GridSearchCV(gbc, param_test, cv=5, refit=True, verbose=1)
    
11. gbc_grid.fit(X_train,y_train)
    
12. train_score = gbc.score(X_train,y_train)
    
13. gbc_pre = gbc.predict(X_valid)
    
14. valid_score = accuracy_score(y_valid,gbc_pre)
    
15. print(valid_score)

复制代码

报错如下

1. Fitting 5 folds for each of 40 candidates, totalling 200 fits
   
2. /opt/conda/lib/python3.10/site-packages/sklearn/base.py:432: UserWarning: X has feature names, but GradientBoostingClassifier was fitted without feature names
   
3.   warnings.warn(
   
4. ---------------------------------------------------------------------------
   
5. NotFittedError                            Traceback (most recent call last)
   
6. Cell In[25], line 12
   
7.      10 gbc_grid = GridSearchCV(gbc, param_test, cv=5, refit=True, verbose=1)
   
8.      11 gbc_grid.fit(X_train,y_train)
   
9. ---> 12 train_score = gbc.score(X_train,y_train)
   
10.      13 gbc_pre = gbc.predict(X_valid)
    
11.      14 valid_score = accuracy_score(y_valid,gbc_pre)
    
12. 
    
13. File /opt/conda/lib/python3.10/site-packages/sklearn/base.py:668, in ClassifierMixin.score(self, X, y, sample_weight)
    
14.     643 """
    
15.     644 Return the mean accuracy on the given test data and labels.
    
16.     645
    
17.    (...)
    
18.     664     Mean accuracy of ``self.predict(X)`` w.r.t. `y`.
    
19.     665 """
    
20.     666 from .metrics import accuracy_score
    
21. --> 668 return accuracy_score(y, self.predict(X), sample_weight=sample_weight)
    
22. 
    
23. File /opt/conda/lib/python3.10/site-packages/sklearn/ensemble/_gb.py:1308, in GradientBoostingClassifier.predict(self, X)
    
24.    1293 def predict(self, X):
    
25.    1294     """Predict class for X.
    
26.    1295
    
27.    1296     Parameters
    
28.    (...)
    
29.    1306         The predicted values.
    
30.    1307     """
    
31. -> 1308     raw_predictions = self.decision_function(X)
    
32.    1309     encoded_labels = self._loss._raw_prediction_to_decision(raw_predictions)
    
33.    1310     return self.classes_.take(encoded_labels, axis=0)
    
34. 
    
35. File /opt/conda/lib/python3.10/site-packages/sklearn/ensemble/_gb.py:1264, in GradientBoostingClassifier.decision_function(self, X)
    
36.    1243 """Compute the decision function of ``X``.
    
37.    1244
    
38.    1245 Parameters
    
39.    (...)
    
40.    1259     array of shape (n_samples,).
    
41.    1260 """
    
42.    1261 X = self._validate_data(
    
43.    1262     X, dtype=DTYPE, order="C", accept_sparse="csr", reset=False
    
44.    1263 )
    
45. -> 1264 raw_predictions = self._raw_predict(X)
    
46.    1265 if raw_predictions.shape[1] == 1:
    
47.    1266     return raw_predictions.ravel()
    
48. 
    
49. File /opt/conda/lib/python3.10/site-packages/sklearn/ensemble/_gb.py:687, in BaseGradientBoosting._raw_predict(self, X)
    
50.     685 def _raw_predict(self, X):
    
51.     686     """Return the sum of the trees raw predictions (+ init estimator)."""
    
52. --> 687     raw_predictions = self._raw_predict_init(X)
    
53.     688     predict_stages(self.estimators_, X, self.learning_rate, raw_predictions)
    
54.     689     return raw_predictions
    
55. 
    
56. File /opt/conda/lib/python3.10/site-packages/sklearn/ensemble/_gb.py:673, in BaseGradientBoosting._raw_predict_init(self, X)
    
57.     671 def _raw_predict_init(self, X):
    
58.     672     """Check input and compute raw predictions of the init estimator."""
    
59. --> 673     self._check_initialized()
    
60.     674     X = self.estimators_[0, 0]._validate_X_predict(X, check_input=True)
    
61.     675     if self.init_ == "zero":
    
62. 
    
63. File /opt/conda/lib/python3.10/site-packages/sklearn/ensemble/_gb.py:382, in BaseGradientBoosting._check_initialized(self)
    
64.     380 def _check_initialized(self):
    
65.     381     """Check that the estimator is initialized, raising an error if not."""
    
66. --> 382     check_is_fitted(self)
    
67. 
    
68. File /opt/conda/lib/python3.10/site-packages/sklearn/utils/validation.py:1390, in check_is_fitted(estimator, attributes, msg, all_or_any)
    
69.    1385     fitted = [
    
70.    1386         v for v in vars(estimator) if v.endswith("_") and not v.startswith("__")
    
71.    1387     ]
    
72.    1389 if not fitted:
    
73. -> 1390     raise NotFittedError(msg % {"name": type(estimator).__name__})
    
74. 
    
75. NotFittedError: This GradientBoostingClassifier instance is not fitted yet. Call 'fit' with appropriate arguments before using this estimator.

复制代码

FishC

从报错信息来看，`GradientBoostingClassifier`实例`gbc`没有经过训练，所以无法进行评分或者预测。这是因为你在`GridSearchCV`中训练的是`gbc_grid`，而你试图用`gbc`进行评分和预测。

以下是你需要做的更改：

1. 使用`gbc_grid.best_estimator_`来获取经过训练的最佳模型实例。

2. 将`gbc.score(X_train, y_train)`改为`gbc_grid.score(X_train, y_train)`。

3. 将`gbc.predict(X_valid)`改为`gbc_grid.predict(X_valid)`。

修改后的代码应该如下所示：

1. param_test = {'n_estimators': [100, 200],
   
2.               'min_samples_split': [2, 5],
   
3.               'max_depth': [3, 4]}
   
4. gbc_grid = GridSearchCV(gbc, param_test, cv=5, refit=True, verbose=1)
   
5. 
   
6. # Train the model using X_train and y_train
   
7. gbc_grid.fit(X_train, y_train)
   
8. 
   
9. # Use the best estimator found
   
10. train_score = gbc_grid.score(X_train, y_train)
    
11. gbc_pre = gbc_grid.predict(X_valid)
    
12. valid_score = accuracy_score(y_valid, gbc_pre)
    
13. 
    
14. print(valid_score)

复制代码

请注意，`param_test`中的参数列表应该包含具体的值供GridSearchCV搜索。另外，确保你传入的`X_train`和`X_valid`数据形式与模型的期望一致。

以上回复来自 FishC AI 小助理，如未能正确解答您的问题，请继续追问。