Code for medium article: An income back test of three roulette strategies based on Reinforcement Learning
The environment setting was based on this repository
play2.0.py -> generate game record
r3.py -> train RL agent
r3-test.py -> test performance
TensorFlow 2.0