4. Machine Learning

Regularization — Quiz

Test your understanding of regularization with 5 practice questions.

Practice Questions

Question 1

When comparing L1 and L2 regularization, which of the following statements most accurately describes their geometric interpretations in terms of the constraint regions for coefficient shrinkage?

Question 2

Consider a scenario where a business analyst is building a predictive model and observes that the model's performance on the validation set is significantly worse than on the training set. This indicates overfitting. If the analyst decides to apply regularization, which of the following is the most direct mechanism by which regularization helps to mitigate this overfitting?

Question 3

In the context of a linear regression model, if the regularization parameter $ \lambda $ is set to an extremely large value, what is the expected outcome for the model's coefficients and its overall performance?

Question 4

A business analyst is tasked with building a predictive model for customer segmentation. The dataset contains a mix of highly relevant features and many noisy, irrelevant features. The analyst wants a model that can automatically identify and select only the most important features, effectively setting the coefficients of irrelevant features to zero. Which regularization technique is best suited for this specific requirement?

Question 5

In a linear regression model with L1 regularization, if two features are highly correlated, what is the typical behavior of their coefficients compared to L2 regularization?