New answers tagged markov-decision-process
1
vote
$E_{\pi}[R_{t+1}|S_t=s,A_t=a] = E[R_{t+1}|S_t=s,A_t=a]$?
Question: Can I write it without the subscript? So $$E_{\pi}[R_{t+1}|S_t=s,A_t=a] = E[R_{t+1}|S_t=s,A_t=a]$$
Yes, your reasoning is sound, there is no need to condition the expectation on the policy, ...
0
votes
How is the state-visitation frequency computed in "Maximum Entropy Inverse Reinforcement Learning"?
Please look at line 5:
If $P(a_{i,j}|s_i)$ is equal to the policy that is used for generating the demonstrated trajectories, then it could be the same. However, in inverse RL you don't know $P(a_{i,j}|...
Top 50 recent answers are included
Related Tags
markov-decision-process × 163reinforcement-learning × 143
pomdp × 19
q-learning × 17
policies × 14
rewards × 12
reward-functions × 12
markov-property × 10
comparison × 9
deep-rl × 9
state-spaces × 9
definitions × 8
value-functions × 8
transition-model × 8
proofs × 7
value-iteration × 7
discount-factor × 7
multi-armed-bandits × 7
environment × 6
machine-learning × 5
reference-request × 5
ai-design × 5
monte-carlo-methods × 5
semi-mdp × 5
policy-gradients × 4