Contributions for exercises

[faizankshaikh]

Inspired from @dennybritz 's RL exercises I want to contribute to building exercises for the DL book.

But before directly contributing here, I would first get feedback from the online deep learning reading group I am part of. Because I would try to make as "non-mediocre" contribution as possible.

Let me know if this is ok with you. Also if you could review the contributions we make, we would be eternally grateful.

Regards,
Faizan

PS: Here's the repo. We've just started so there's not much there.