ConceptReward Shaping Invariance
Reward Shaping Invariance
A theorem property proved by Ng et al. in 1999: if the shaping reward takes the form F(s,s')=γΦ(s')−Φ(s), then the optimal policy obtained under this shaping is exactly the same as the optimal policy for the original sparse-reward problem, meaning potential-based shaping does not change the optimal policy.
Timeline (last 90 days)
Oct 14
If the shaping reward takes the form F(s, s') = γΦ(s') − Φ(s), then the optimal policy obtained under this shaping is exactly identical to the optimal policy of the original sparse-reward problem.
Unverified50%