A2C: An actor-critic DRL algorithm that uses a neural network to approximate the policy and the value function and parallel workers to sample the environment