👉 Ridge regression is a type of linear regression that adds an L2 regularization term to the loss function, which helps prevent overfitting by penalizing large coefficients. This regularization encourages simpler models by shrinking the magnitude of the coefficients towards zero, but not exactly to zero, unlike Lasso regression. The ridge penalty term is proportional to the square of the magnitude of the coefficients, making it effective in handling multicollinearity and improving model generalization. By balancing the trade-off between fitting the training data well and keeping the model coefficients small, ridge regression provides a robust solution for datasets with correlated features or when multicollinearity is a concern.
