wagenaartje / neataptic

:rocket: Blazing fast neuro-evolution & backpropagation for the browser and Node.js
https://wagenaartje.github.io/neataptic/
Other
1.19k stars 279 forks source link

How i made reinforcement learning ? #122

Open pks1989 opened 6 years ago

pks1989 commented 6 years ago

Hi!

First, thank you for this project ! :+1:

Next, i try use this project teach neuronet play poker. I was understood all aspects of architecture future program, except one: how i may do reinforcement learning ?

I dont now, how win poker game, but i have transparent assessment tool of success - size of the player's bank.

I want do 4 neuronetworks for make decision at each stage of poker game and and evaluate the result :)