Hi,
Here are some questions I feel confused about:
1. Last year exam Yes-No questions: 2. (can you give me an example where learning rate is a function of the model output); 3. ; 8.; 13.; 15.(i think this false according to the no free lunch theorem, but i have no idea how to reconcile the no free lunch theorem with the statement that deep learning can fit anything given enough parameters; do they contradict?)
2. Last year exam 6. Policy gradient: could you go through the solution for us? Especially for the first two sub-questions
3. Exercise set 11: Exercise 2, c)
4. Exercise set 11: Exercise 5. d)
Thank you in advance!
