DecodeAI
← Question Bank

Machine Learning

Linear Methods For Regression

Interview questions on Linear Methods For Regression.

33 questions

Shrinkage Methods

Q2. What are the benefits of using shrinkage methods over subset selection methods?

Sign in to bookmark

Shrinkage Methods

Q4. What's the main purpose of L1 and L2 regularization in linear regression?

Sign in to bookmark

Shrinkage Methods

Q5. How do we estimate coefficients in Ridge regression?

Sign in to bookmark

Shrinkage Methods

Q6. Explain the effect of tuning parameter $\lambda$ in ridge regression cost function?

Sign in to bookmark

Shrinkage Methods

Q7. How can we determine optimal value of $\lambda$ ?

Sign in to bookmark

Shrinkage Methods

Q8. Suppose you fit a ordinary linear regression model over your data and you find it is under-fitting. Is it good idea to use Ridge or Lasso regression here?

Sign in to bookmark

Shrinkage Methods

Q9. How do L1 and L2 regularization affect the model's coefficients?

Sign in to bookmark

Shrinkage Methods

Q10. Can we use ridge regression for variable selection purpose?

Sign in to bookmark

Shrinkage Methods

Q11. Write the loss function involve in lasso regression?

Sign in to bookmark

Shrinkage Methods

Q12. Which regression technique leads to sparse models? Lasso or ridge regression?

Sign in to bookmark

Shrinkage Methods

Q13. Why is it that the lasso, unlike ridge regression, results in coefficient estimates that are exactly equal to zero?

Sign in to bookmark

Shrinkage Methods

Q14. List some advantages of using lasso regression over ridge regression?

Sign in to bookmark

Shrinkage Methods

Q15. What are the hyper-parameters associated with L1 and L2 regularization?

Sign in to bookmark

Shrinkage Methods

Q16. When would you choose L1 regularization over L2, and vice versa?

Sign in to bookmark

Shrinkage Methods

Q17. What is Elastic Net regularization, and how does it relate to L1 and L2 regularization?

Sign in to bookmark

Shrinkage Methods

Q18. How do you choose the optimal regularization strength (alpha) for L1 and L2 regularization?

Sign in to bookmark

Shrinkage Methods

Q19. What are the consequences of multi-collinearity?

Sign in to bookmark

Shrinkage Methods

Q20. How can you detect collinearity in a regression model?

Sign in to bookmark

Shrinkage Methods

Q21. How can you detect multi-collinearity in a regression model?

Sign in to bookmark

Shrinkage Methods

Q24. How can we determine whether a dataset is high-dimensional or low-dimensional?

Sign in to bookmark

Shrinkage Methods

Q25. What is the issue of using least squares regression in high dimensional setting?

Sign in to bookmark

Shrinkage Methods

Q26. What happens to the Train MSE and Test MSE in a linear regression model if we add features that are completely unrelated to the response?

Sign in to bookmark

Shrinkage Methods

Q28. If I add a feature to my linear regression model, how will it affect the Train MSE and Test MSE?

Sign in to bookmark

Shrinkage Methods

Q29. Can we use lasso or ridge regression in high dimensional setup?

Sign in to bookmark

Shrinkage Methods

Q30. Can we use lasso or ridge regression in high dimensional setup? (Part 2)

Sign in to bookmark

Shrinkage Methods

Q31. [True/False] In the high-dimensional setting, the multicollinearity problem is extreme?

Sign in to bookmark

Shrinkage Methods

Q32. [True/False] In the high-dimensional setting, the multicollinearity problem is extreme? (Part 2)

Sign in to bookmark

Shrinkage Methods

Q33. Why should one be cautious when reporting errors and measures of model fit in high-dimensional settings?

Sign in to bookmark