sklearn 源码解析基本线性模型岭回归 ridge.py(2)

斯温jack

于 2016-08-19 05:59:27 发布

阅读量1.5k

点赞数

分类专栏：机器学习 Sklearn 文章标签： scikit-learn sklearn 岭回归机器学习 ridge

本文链接：https://blog.youkuaiyun.com/sinat_30665603/article/details/52247080

版权

本文详细解析了sklearn中Ridge回归的源码实现，包括_BaseRidge类的数据预处理、fit方法的实现，以及Ridge和RidgeClassifier两个子类的功能和区别。Ridge回归主要处理线性模型，而RidgeClassifier则适用于二分类问题。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

class _BaseRidge(six.with_metaclass(ABCMeta, LinearModel)):

 @abstractmethod
 def __init__(self, alpha = 1.0, fit_intercept = True, normalize = False, 
    copy_X = True, max_iter = None, tol = 1e-3, solver = "auto", random_state = None):
  self.alpha = alpha 
  self.fit_intercept = fit_intercept
  self.normalize = normalize 
  self.copy_X = copy_X 
  self.max_iter = max_iter 
  self.solver = solver
  self.random_state = random_state 

 def fit(self, X, y, sample_weight = None):
  X, y = check_X_y(X, y, ['csr', 'csc', 'coo'], dtype = np.float64, 
      multi_output = True, y_numeric = True)

  if ((sample_weight is not None) and np.atleast_1d(sample_weight).ndim > 1):
   raise ValueError("Sample weights must be 1D array or scalar")

  X, y, X_offset, y_offset, X_scale = self._preprocess_data(X, y, self.fit_interce