Q.3 what is greedy agent? Does a greedy agent always find an optimal policy?
Answers
Answered by
6
Answer:
it can be anything yeah
Answered by
3
Answer:
To try an interactive version, go here. ... This is referred to as a greedy method. Taking the action which the agent estimates to be the best at the current moment is an example of exploitation: the agent is exploiting its current knowledge about the reward structure of the environment to act.
There is always at least one such optimal policy[8]. The so called greedy policy is following the currently best path of actions. During learning however, for the values to converge into good estimates it is required that the agent visits all available states to gain information about them.
Similar questions