Python超参数调优遇NameError：RandomForestClassifier未定义求解

问题原因 #

报错提示NameError: name 'RandomForestClassifier' is not defined，是因为你只导入了RandomizedSearchCV，但没有导入RandomForestClassifier类，同时accuracy_score也需要单独导入。

解决方案 #

1. 补充缺失的导入语句
   在代码开头添加以下导入：

   from sklearn.ensemble import RandomForestClassifier
   from sklearn.metrics import accuracy_score
   

2. 完整可运行代码示例

   from sklearn.ensemble import RandomForestClassifier
   from sklearn.metrics import accuracy_score
   from sklearn.model_selection import RandomizedSearchCV
   
   n_estimators_array = [1, 4, 5, 8, 10, 20, 50, 75, 100, 250, 500]
   results = []
   for n in n_estimators_array:
       forest = RandomForestClassifier(n_estimators=n, random_state=76)
       forest.fit(x_train, y_train)
       result = accuracy_score(y_test, forest.predict(x_test))
       results.append(result) 
       print(n, ':', result)
   

额外优化（用RandomizedSearchCV做超参数调优） #

既然你已经导入了RandomizedSearchCV，可以直接用它替代手动循环，效率更高且支持交叉验证：

from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import accuracy_score
from sklearn.model_selection import RandomizedSearchCV

param_dist = {'n_estimators': [1, 4, 5, 8, 10, 20, 50, 75, 100, 250, 500]}
# 初始化随机搜索，cv=5表示5折交叉验证
random_search = RandomizedSearchCV(
    estimator=RandomForestClassifier(random_state=76),
    param_distributions=param_dist,
    n_iter=11,  # 遍历所有参数选项
    cv=5,
    random_state=76,
    scoring='accuracy'
)
random_search.fit(x_train, y_train)

# 输出最佳结果
print("最佳n_estimators参数:", random_search.best_params_)
print("交叉验证最佳准确率:", random_search.best_score_)
# 在测试集上验证
test_accuracy = accuracy_score(y_test, random_search.predict(y_test))
print("测试集准确率:", test_accuracy)

内容的提问来源于stack exchange，提问作者Kori gray