In one of the Quiz below we have more then 2 actions which we generally used inside lecture.
You are controlling a robot that has 3 actions: ← (left), → (right) and STOP. From a given state.
s, you have computed Q(s, ←) = -10, Q(s, →) = -20, Q(s, STOP) = 0.
What is the optimal action to take in state?
In this case over all you will be having two terminal states but you can do three actions to go either side but will follow where rewards are higher as the basic rule of thumb - is my understanding right?
Hi @Nitin_Gautam,
welcome to the community!
In the outlined example you are right: there are three possible actions which will lead to a different reward (which should be maximised overall according to your optimization task.)
Here some more info which might be useful to read.
Best regards
Christian
Hi @Nitin_Gautam,
Yes but please note, that in many situations, always going for higher reward actions may not be the most efficient or effective way to achieve a goal. While high reward actions may offer the potential for greater rewards, they may also come with greater risks or longer time horizons, which can lead to delays or setbacks.
In some cases, it may be more advantageous to take optimal actions that have a fair reward but also take less time to reach the terminal state. This can help to ensure that progress is being made consistently, and that the overall goal is achieved in a timely manner.