Multiple Choice Questions
What is the primary purpose of Lasso, Ridge, and Elastic Net regularization techniques?
A) To increase the complexity of the model
B) To prevent overfitting by penalizing large coefficients
C) To reduce the number of features
D) To increase the number of features
Show answer
Answer: B) To prevent overfitting by penalizing large coefficients
Ridge regression adds a penalty equal to which of the following?
A) The absolute value of the coefficients
B) The square of the coefficients
C) The cube of the coefficients
D) The logarithm of the coefficients
Show answer
Answer: B) The square of the coefficients
Elastic Net regularization combines which two types of penalties?
A) L1 and L3 penalties
B) L2 and L3 penalties
C) L1 and L2 penalties
D) None of the above
Show answer
Answer: C) L1 and L2 penalties
True/False Questions
True or False: Lasso regression tends to produce sparse models by eliminating some features.
Show answer
Answer: True
True or False: Ridge regression can perform feature selection by setting some coefficients to zero.
Show answer
Answer: False (Ridge regression shrinks coefficients but does not set them to zero.)
True or False: Increasing the regularization parameter (?) in Ridge regression decreases the magnitude of the coefficients.
Show answer
Answer: True
True or False: Lasso regression is more robust to outliers than Ridge regression.
Show answer
Answer: False (Lasso is less robust to outliers because it uses L1 norm which is more sensitive to large errors compared to L2 norm used by Ridge regression.)
Discover more from Science Comics
Subscribe to get the latest posts sent to your email.