Near-Optimal Region
In simulation optimization, the set of designs whose expected performance falls within a tolerance range of the optimal value, used to support multi-criteria decision making.
Timeline (last 90 days)
What decision-makers really need is the set of designs whose expected performance falls within some tolerance of the optimum, i.e., the near-optimal region.
The paper studies how to simultaneously identify the near-optimal regions for all tolerances (up to a given upper limit T) in a continuous design space while guaranteeing that the region estimation error does not exceed η.
When α > ν, the lower bound on the minimum number of simulation evaluations required by any algorithm to simultaneously identify the near-optimal regions for all tolerances is Ω(η^{-(2+d/ν)} min{1, KT^{d/α}}).
All Facts (3)
What decision-makers really need is the set of designs whose expected performance falls within some tolerance of the optimum, i.e., the near-optimal region.
50%UnverifiedThe paper studies how to simultaneously identify the near-optimal regions for all tolerances (up to a given upper limit T) in a continuous design space while guaranteeing that the region estimation error does not exceed η.
50%UnverifiedWhen α > ν, the lower bound on the minimum number of simulation evaluations required by any algorithm to simultaneously identify the near-optimal regions for all tolerances is Ω(η^{-(2+d/ν)} min{1, KT^{d/α}}).
50%