Request edit access
How to improve practical RL
Hello!
We've recently finished the first large iteration of practical_RL course. It would be fair to say that you contributed least just as much work as we did: extensive feedback, contributed bugfixes, recipes for running in the cloud/via docker, compatibility patches, tensorflow versions of the assignments - together we accomplished much more than "official" course staff could ever have done.

And yet there's still much to improve. It is very likely (unless majority votes against that below) the course will be repeated next year with newblood students. We need your opinion on how to improve course for them.

TL;DR please give us feedback to make the course more productive for the next generation of YSDA&HSE students. Your feedback is especially important if you found something unsatisfactory.

Note: the survey is to make the course better, not to make us feel good. In case of doubt, please choose the worse option (idk if "good" or "okay" -> "okay",  "okay" or "bad" -> "bad").
Note: please keep your reply anonymous.
Sign in to Google to save your progress. Learn more
General impression
Here you basically decide the course fate and state your overall impression.
Worth your time? *
Required
Technical aspects *
Severe problems
Some problems
Okay
Exceeding expectations
Technical stuff: installing libraries, docker, etc
Technical stuff: anytask, homework submission
Lectures in general
Seminar & HW assignments
Seminar & HW checkup
Q&A & feedback
Course materials (optional)
Please tell us your impression about each week
week0: intro & evolution
week1: crossentropy method
week2: TD & q-learning
week3: value-based
week3.5: intro to DL & theano (no lecture)
week4: approximate rl
week5: deep rl
week6: policy gradients
week7: partially observable mdp
week8: seq2seq; translation
week9.2; exploration (by yozhik)
week9.1; exploration (by alvor88)
week10: TRPO
week11: planning; MCTS
week12: other cool stuff
Your suggestions for course improvement
Please don't hesitate to write anything we should know to improve:  1. Course organization  2. Practice assignments   3. Lectures structure and content  4.  Personal  material presentation skills
Submit
Clear form
Never submit passwords through Google Forms.
This content is neither created nor endorsed by Google. - Terms of Service - Privacy Policy

Does this form look suspicious? Report