https://gist.github.com/viswanathgs/ca9788020cfcc7849b9181d9239a2ef4

https://medium.com/@tuzzer/follow-up-cart-pole-balancing-with-q-network-976d13f88d2f

https://github.com/Microsoft/CNTK/blob/9231ece8cec2fa25b4ca06018854cfe3ffb8ea53/Tutorials/CNTK_203_Reinforcement_Learning_Basics.ipynb