This project is an attempt at creating an agent that can play Texas Hold'em poker at a good level. Agents use neural networks to decide what to do at each point in time, and can be trained with this program. Currently elitist and genetic evolutions are possible, in order to train agents.
The agents currently evolve to realize that folding is the best way to success, since they are eliminating losses completely... But work is being done to try to get them to play more confidently, without resorting to constant folding.
More information will be added as the project progresses, but since it is currently in experimental phase, too much is changing to fully document or explain. Come back later to see how the project is evolving!