3 ms·
Q* is the optimal (ie, correct for the decision problem) function computing the total expected reward of taking an action from a given state in reinforcement le
by sudosysgen 3y ago
Q* is the optimal (ie, correct for the decision problem) function computing the total expected reward of taking an action from a given state in reinforcement learning.
- NhanH 3y agoThat is just Q. The asterisk is new
- debugnik 3y agoThe "star" in A* means optimal (as in actually proven to be optimal, so they could stop publishing A^[whatever] algorithms). I assume either Q* is considered optimal in a way regular Q-learning isn't, or they're mixing Q-learning with some A*-like search algorithm. (Or someone picked an undeserving name.)
- sudosysgen 3y agoDepends on the notation. Sometimes Q* is used to denote optimality, for example here : https://www.cs.toronto.edu/~jlucas/teaching/csc411/lectures/lec21_22_handout.pdf https://www.cs.toronto.edu/~jlucas/teaching/csc411/lectures/... , page 31