Python 逻辑回归错误：“TypeError: issubclass() arg 2 must be a class or tuple of classes”

Question

我正在创建一个具有 4 种可能结果的多类分类模型。它昨天工作，但今天，我收到以下错误。我对 Python 不是很熟悉，因此我们非常感谢有关如何解决此问题的任何帮助。

from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)

from sklearn.preprocessing import StandardScaler

scaler = StandardScaler()
scaler.fit(X_train)
X_train_scaled = scaler.transform(X_train)
X_test_scaled = scaler.transform(X_test)

Logistic=LogisticRegression()
logistic.fit(X_train_scaled,y_train)
y_pred_log=logistic.predict(X_test_scaled)
log_cm=(metrics.confusion_matrix(y_test, y_pred_log))

TypeError                                 Traceback (most recent call last)
<ipython-input-213-8e436855d9cc> in <module>
      1 logistic=LogisticRegression()
----> 2 logistic.fit(X_train_scaled,y_train)
      3 y_pred_log=logistic.predict(X_test_scaled)
      4 log_cm=(metrics.confusion_matrix(y_test, y_pred_log))

~\Anaconda3\lib\site-packages\sklearn\linear_model\logistic.py in fit(self, X, y, sample_weight)
   1491         The SAGA solver supports both float64 and float32 bit arrays.
   1492         """
-> 1493         solver = _check_solver(self.solver, self.penalty, self.dual)
   1494 
   1495         if not isinstance(self.C, numbers.Number) or self.C < 0:

~\Anaconda3\lib\site-packages\sklearn\linear_model\logistic.py in _check_solver(solver, penalty, dual)
    430         warnings.warn("Default solver will be changed to 'lbfgs' in 0.22. "
    431                       "Specify a solver to silence this warning.",
--> 432                       FutureWarning)
    433 
    434     all_solvers = ['liblinear', 'newton-cg', 'lbfgs', 'sag', 'saga']

TypeError: issubclass() arg 2 must be a class or tuple of classes

Answer 1

似乎您的求解器导致错误。尝试更改您的求解器：

solver = 'lbfgs'