Optimal hysteresis for a class of deterministic deteriorating two-armed bandit problem with switching costs. (Q1421425): Difference between revisions

The authors derive the optimal policy for the dynamic scheduling of a class of special two-armed bandit-problems with switching costs. The scheduling policy exhibits an hysteretic character. Using this solvable class of models, the authors observe the performance of a sub-optimal policy derived from a set of generalized priority indices.