cs285-lec7-value function
If we don't want to learn a policy, we could learn value funtions.
0%
If we don't want to learn a policy, we could learn value funtions.
To deal with high-variance in policy gradient, we introduce actor-critic