We consider the infinite version of the secretary problem. From an infinite stream of applicants, we are to select m choices and assign them, in order, to m positions to be filled. We receive a reward c k (m) for 1 ≤ k ≤ m(0 ≤ c 1 (m) ≤ ⋯ ≤ c m (m) = 1) if the best k applicants are selected and correctly ranked (i.e., the ith best applicant is assigned to the ith position for 1 ≤ i ≤ k). We consider two versions of the problem: (i) once an applicant is selected and assigned to some position, he cannot be promoted to a better position at a later time, and (ii) previously selected applicants can be promoted. In this paper, we concentrate on the double selection problem, m = 2, and derive, for both the promotion and nonpromotion versions of the problem, the optimal strategy that maximizes the expected reward.
No takes yet. Share an insight, caveat, or question.
Mitsushi Tamaki (1986) studied this question.
Synapse has enriched 4 closely related papers on similar clinical questions. Consider them for comparative context: