On a connection between total positivity and last-success stopping problems

Consider the optimal stopping problem of maximising the expected payoff from selecting the last success in a sequence of independent Bernoulli trials. The total positivity of the Markov chain embedded in the success epochs of the trials is exploited to prove the optimality of the myopic strategy for...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
1. Verfasser: Derbazi, Zakaria
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Consider the optimal stopping problem of maximising the expected payoff from selecting the last success in a sequence of independent Bernoulli trials. The total positivity of the Markov chain embedded in the success epochs of the trials is exploited to prove the optimality of the myopic strategy for both unimodal stopping and continuation payoffs. In contrast, the problem is shown to be nonmonotone for oscillating payoffs, alternating between two values. The myopic strategy is demonstrated not to be a threshold rule. Lastly, the optimality of the myopic policy is established for the $\ell$th-to-$m$th last-success problem based on a total positivity argument. An illustrative example is given to compute the asymptotic threshold and winning probability.
DOI:10.48550/arxiv.2411.07103