Deep Q Learning Implemented for dl4j. Code is as-is, sorry its not commented, and its probably not standard java convention
TestRL is a sample pathfinding problem which seems to work for me. See how that is set up to use the DeepQNetork
The test problem was inspired from this blog tutorial: http://outlace.com/Reinforcement-Learning-Part-3/