The CAA exists in two environments (one is the behavioural environment where it behaves), and the other is the genetic environment, wherefrom it initially and only once receives initial emotions about situations to be encountered in the behavioural environment. There is neither a separate reinforcement input nor an advice input from the environment. Reinforcement learning algorithms do not assume knowledge of an exact mathematical model of the MDP and are used when exact models are infeasible.
Test Post Created
Test Post Created