Assignment 2 Q11

Hi, my question is from q11 in the 2nd assignment, .

After working further on the problem, you’ve decided to correct the incorrectly labeled data on the dev set. Which of these statements do you agree with? (Check all that apply).

My question is

  1. How does fixing incorrectly label data in the training set change its distribution?
  2. Why is it the DL algorithms are robust to having slightly different train and dev distributions?

Thank you!

Hi @shamus , same as in your other post, please remove direct references from the quizzes. We can discuss about any concepts, but without direct answers to the quizzes

So sorry for that. I changed the question already.