本文件根据../textclf/config/ml_model.py自动生成
无可设置的属性
参考:https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html
LogisticRegressionConfig继承MLModelConfig的所有属性，同时它还有以下属性：
| Attribute name | Type | Default | Description |
|---|---|---|---|
| penalty | str | "l2" | penalty{‘l1’, ‘l2’, ‘elasticnet’, ‘none’} Used to specify the norm used in the penalization. |
| dual | bool | False | Dual or primal formulation. Dual formulation is only implementedfor l2 penalty with liblinear solver. Prefer dual=False when n_samples > n_features. |
| tol | float | 1e-4 | Tolerance for stopping criteria. |
| C | float | 1.0 | Inverse of regularization strengthmust be a positive float. Like in support vector machines, smaller values specify stronger regularization. |
| fit_intercept | bool | True | Specifies if a constant(a.k.a. bias or intercept) should be added to the decision function. |
| intercept_scaling | float | 1 | Useful only when the solver ‘liblinear’ is used and self.fit_intercept is set to True. |
无可设置的属性