December 2018
Intermediate to advanced
764 pages
18h 18m
English
If the model is not available then the agent learns the model and optimal policy by trial and error. When the model is not available, the agent uses a Q function, which is defined as follows:

The Q function basically maps the pairs of states and actions to a real number that denotes the expected total reward if the agent at state s selects an action a.
Read now
Unlock full access