State driven safety for reinforcement learning without reward
This is an implementation of different classical Reinforcement Learning algorithms without reward function. The training is State Goal driven and focus on safety.
Actual implemented algorithms: DQN