I can't seem to understand the difference between q-learning and utility learning clearly, I know that q learning means learning the utility of actions available to a state rather than learning the utility of each state(util learning). But when it comes to its applications, I don't get which one is to be used for any case. Maybe if explained with an example, that might help