Deep Q-learning

Deep Q-learning represents an evolution of the basic Q-learning method the state-action is replaced by a neural network, with the aim of approximating the optimal value function.

Compared to the previous approaches, where it was used to structure the network in order to request both input and action and providing its expected return, Deep Q-learning revolutionizes the structure in order to request only the state of the environment and supply as many status-action values as there are actions that can be performed in the environment.