Standardizing the documentation for `alpha` in `Ridge` and `Lasso`

Describe the issue linked to the documentation

I’m helping to teach a machine learning course using sklearn. The documentation for alpha in Ridge is as follows:

Regularization strength; must be a positive float. Regularization improves the conditioning of the problem and reduces the variance of the estimates. Larger values specify stronger regularization. Alpha corresponds to 1 / (2C) in other linear models such as LogisticRegression or sklearn.svm.LinearSVC. If an array is passed, penalties are assumed to be specific to the targets. Hence they must correspond in number.

The documentation for alpha in Lasso is as follows

Constant that multiplies the L1 term. Defaults to 1.0. alpha = 0 is equivalent to an ordinary least square, solved by the LinearRegression object. For numerical reasons, using alpha = 0 with the Lasso object is not advised. Given this, you should use the LinearRegression object.

Between the different documentation for alpha in these models, plus the C being the inverse regularization in the closely related LogisticRegression there has understandably been some confusion for myself and some students.

Suggest a potential alternative/fix

The two models are closely related, and I think standardizing the description of alpha to be the same in Ridge and Lasso (making appropriate L2/L1 changes respectively) would go a long way to preventing some confusion students have been encountering. The documentation for alpha in Lasso seems to be most straightforward to me so I would propose using that description in both Lasso and Ridge.

[TomDLT]

The API page for each class is automatically generated from the docstring: https://github.com/scikit-learn/scikit-learn/blob/master/sklearn/linear_model/_ridge.py#L603-L617 https://github.com/scikit-learn/scikit-learn/blob/master/sklearn/linear_model/_coordinate_descent.py#L904-L915

[Rubix982]

@Dpananos please take a look at the PR. Thanks!