Paperlane

no. 3 · Tuesday 27 October 2026

On optimal stopping problems with full information and rank-dependent reward

Alexander Goldenshluger and Yaakov Malinovsky

math.OC

Abstract

In this paper we consider optimal stopping problems with full information and rank-dependent rewards. Our focus is on the following two problems: selecting one of the r best observations (Gusein-Zade's problem with full information) and minimizing the expected rank (Robbins' problem). In both settings we provide full characterization of the optimal stopping rules and corresponding optimal values. We also present numerical experiments in which the derived dynamic programming recursions are implemented, and approximations to optimal values are computed.

comments
Keywords: optimal stopping, full information, rank-dependent reward, dynamic programming, Robbins' problem, Gusein-Zade's problem
fields
Optimization and Control, Probability, Statistics Theory
pdf ↗ page arXiv:2610.33701 ↗