为什么在 sklearn GridsearchCV SVM 中使用 class_weight 时会出现错误？

Question

下面是我的代码：

tuned_parameters = [
    {'kernel': ['linear], 'C':[1, 10], 'class_weight': ['auto']}, {'kernel': ['rbf'], 'C':[1,10], 'class_weight':['auto']}]
clf = GridSearchCV(svm.SVC(), tuned_parameters, cv=5, scoring='accuracy')
clf.fit(x_train,y_train)

但是我收到以下错误：

Traceback (most recent call last):
  File "/home/arajabi/PycharmProjects/Muffin/classification.py", line 77, in <module>
    clf3.fit(x_train, y_train)
  File "/home/arajabi/anaconda3/lib/python3.5/site-packages/sklearn/model_selection/_search.py", line 639, in fit
    cv.split(X, y, groups)))
  File "/home/arajabi/anaconda3/lib/python3.5/site-packages/sklearn/externals/joblib/parallel.py", line 779, in __call__
    while self.dispatch_one_batch(iterator):
  File "/home/arajabi/anaconda3/lib/python3.5/site-packages/sklearn/externals/joblib/parallel.py", line 625, in dispatch_one_batch
    self._dispatch(tasks)
  File "/home/arajabi/anaconda3/lib/python3.5/site-packages/sklearn/externals/joblib/parallel.py", line 588, in _dispatch
    job = self._backend.apply_async(batch, callback=cb)
  File "/home/arajabi/anaconda3/lib/python3.5/site-packages/sklearn/externals/joblib/_parallel_backends.py", line 111, in apply_async
    result = ImmediateResult(func)
  File "/home/arajabi/anaconda3/lib/python3.5/site-packages/sklearn/externals/joblib/_parallel_backends.py", line 332, in __init__
    self.results = batch()
  File "/home/arajabi/anaconda3/lib/python3.5/site-packages/sklearn/externals/joblib/parallel.py", line 131, in __call__
    return [func(*args, **kwargs) for func, args, kwargs in self.items]
  File "/home/arajabi/anaconda3/lib/python3.5/site-packages/sklearn/externals/joblib/parallel.py", line 131, in <listcomp>
    return [func(*args, **kwargs) for func, args, kwargs in self.items]
  File "/home/arajabi/anaconda3/lib/python3.5/site-packages/sklearn/model_selection/_validation.py", line 458, in _fit_and_score
    estimator.fit(X_train, y_train, **fit_params)
  File "/home/arajabi/anaconda3/lib/python3.5/site-packages/sklearn/svm/base.py", line 150, in fit
    y = self._validate_targets(y)
  File "/home/arajabi/anaconda3/lib/python3.5/site-packages/sklearn/svm/base.py", line 502, in _validate_targets
    self.class_weight_ = compute_class_weight(self.class_weight, cls, y_)
  File "/home/arajabi/anaconda3/lib/python3.5/site-packages/sklearn/utils/class_weight.py", line 62, in compute_class_weight
    " got: %r" % class_weight)
ValueError: class_weight must be dict, 'balanced', or None, got: 'auto'

我对 python 比较陌生。有人可以给我一个简单的解决方案吗？

Answer 1

您为 SVC 传入了 invalid parameter、'auto'。错误消息会告诉您可能的值是什么：classes、balanced 或 None 的字典。如果您希望它使用默认行为，请将其留空或输入 None

class_weight : {dict, ‘balanced’}, optional

通常当我对 SVC 进行网格搜索时，我会使用 'class_weight': [None, 'Balanced] 除非我有一个特定的 class 余额我想尝试

Answer 2

这不是典型的 Python 错误，GridSearchCV 不喜欢 class_weight 中的 auto 参数：

tuned_parameters = [
{'kernel': ['linear], 'C':[1, 10], 'class_weight': ['auto' <---

我不熟悉，我只能重复错误信息所说的内容：

class_weight must be dict, 'balanced', or None, got: 'auto'

要了解更多信息，您必须查看 http://scikit-learn.org/stable/modules/generated/sklearn.model_selection.GridSearchCV。

Answer 3

sklearn.svm.SVC的文档是对的here。参数 class_weight 不接受 'auto' 作为输入值。那是你的错误。

您可以通过更换解决此问题：

'class_weight': ['auto']

与：

'class_weight': ['balanced']

为什么在 sklearn GridsearchCV SVM 中使用 class_weight 时会出现错误？

Why do I get error when using class_weight in sklearn GridsearchCV SVM?

python

class

svm

weighted

scikit-learn