[source] # Ridge L2 regularized linear regression solved using a closed-form solution. The addition of regularization, controlled by the *alpha* hyper-parameter, makes Ridge less likely to overfit the training data than ordinary least squares (OLS). **Interfaces:** [Estimator](../estimator.md), [Learner](../learner.md), [Ranks Features](../ranks-features.md), [Persistable](../persistable.md) **Data Type Compatibility:** Continuous ## Parameters | # | Name | Default | Type | Description | |---|---|---|---|---| | 1 | l2Penalty | 1.0 | float | The strength of the L2 regularization penalty. | ## Example ```php use Rubix\ML\Regressors\Ridge; $estimator = new Ridge(2.0); ``` ## Additional Methods Return the weights of features in the decision function. ```php public coefficients() : array|null ``` Return the bias added to the decision function. ```php public bias() : float|null ```