This is simply a transpile / reproduction of the original Torch version from Deepmind, but in TensorFlow. It doesn't really do anything new or different compared to the paper by [Mnih et. al.](http://www.nature.com/nature/journal/v518/n7540/full/nature1...) .
Replicating a previous result is still a worthwhile endeavor. Frankly, I don't think the adjective "simply" applies here. This isn't a script kiddie running a binary they found on the web.
P.S. I'm a professional scientist and I think the world would be better if people replicated results more often, and even showcased negative results.
Would be interesting to see if it could be regularized to make it a bit less twitchy. (E.g. by giving a fitness bonus to no action / and-or penalizing many changes of direction within some time window.)
[+] [-] smhx|10 years ago|reply
[+] [-] tostitos1979|10 years ago|reply
P.S. I'm a professional scientist and I think the world would be better if people replicated results more often, and even showcased negative results.
[+] [-] nullc|10 years ago|reply
[+] [-] minimaxir|10 years ago|reply