Solving an RL problem requires an estimate, during the learning process, of an evaluation function. This function must be able to assess, through the sum of the rewards, the success of a policy.
The basic idea of Q-Learning is that the algorithm learns the optimal evaluation function for the entire space of states and actions (S × A). This so-called Q-function provides a match in the form Q: S × A -> R, where R is the expected value of the future rewards of an action 
executed in the state, 
. Once the agent has learned the optimal function, Q, it will be able to recognize what action will lead to the highest future reward in a certain state.
One of the most commonly used examples of implementing the Q-Learning algorithm involves the use of a table. Each cell of the table is a value Q(s; a)= R and it is initialized to 0. The action 
, performed by the agent, is chosen using a policy which is epsilon-greedy with respect to Q.
The basic idea of the Q-Learning algorithm...