SASORI
An adaptive sampling algorithm for continuous simulation optimization that simultaneously identifies near-optimal regions under all tolerances and estimates the optimal point, with sample complexity attaining minimax optimality.
Timeline (last 90 days)
The SASORI (Simultaneous Adaptive Sampling for Optimal Region Identification) algorithm uses the same set of simulation evaluations to simultaneously return an estimated optimum and the near-optimal regions for all tolerances.
The sample complexity of SASORI matches the lower bound up to logarithmic factors and is therefore minimax optimal.
SASORI achieves an optimal trade-off between region estimation error and cumulative regret.
When the tolerance is zero, SASORI reduces to a minimax optimal simulation optimization procedure.
Numerical experiments show that SASORI can accurately identify near-optimal regions and is more sample-efficient than naive methods.
The method is expected in the future to be extended to more complex constraints and higher-dimensional problems.
All Facts (6)
The SASORI (Simultaneous Adaptive Sampling for Optimal Region Identification) algorithm uses the same set of simulation evaluations to simultaneously return an estimated optimum and the near-optimal regions for all tolerances.
50%UnverifiedThe sample complexity of SASORI matches the lower bound up to logarithmic factors and is therefore minimax optimal.
50%UnverifiedSASORI achieves an optimal trade-off between region estimation error and cumulative regret.
50%UnverifiedWhen the tolerance is zero, SASORI reduces to a minimax optimal simulation optimization procedure.
50%UnverifiedNumerical experiments show that SASORI can accurately identify near-optimal regions and is more sample-efficient than naive methods.
50%UnverifiedThe method is expected in the future to be extended to more complex constraints and higher-dimensional problems.
50%