Bellman Equation with Misstep Prob

rmwkwok January 12, 2024, 3:14pm 2

It is the default to move left if they are equal.

Check this out for the result of 0.5 misstep probability.

Can you share a screenshot like in the above link that contains the two tables (Optimal policy and Q(s,a)) for the 0.7 case?

Cheers,
Raymond

Topic		Replies	Views
State Action Value Function misstep_prob = 0.9999 favors low rewards Unsupervised Learning, Recommenders, Reinforcement week-3	1	477	August 10, 2022
State-action value function example question Unsupervised Learning, Recommenders, Reinforcement week-3	5	556	September 27, 2022
Week 3 Reinforcement Learning, Quiz Optional Unsupervised Learning, Recommenders, Reinforcement week-3	4	494	January 3, 2023
Random Stochastic Environment Question Unsupervised Learning, Recommenders, Reinforcement week-3	2	488	August 8, 2022
State-Action value fails to find optimal policy Unsupervised Learning, Recommenders, Reinforcement week-3	2	520	January 29, 2023