On approximative solutions of optimal stopping problems.

*(English)*Zbl 1235.60038In a series of papers (see, e.g., R. Kühne and L. Rüschendorf [Theory Probab. Appl. 48, No. 3, 465–480 (2003) and Teor. Veroyatn. Primen. 48, No. 3, 557–575 (2003; Zbl 1079.60043)]) an approximation method was developed in order to solve approximatively optimal stopping problem for sequences \(X_1,\dots, X_n\) of independent or dependent random variables by some limiting stopping problems for Poisson and related point processes. The basic assumption in this approach is convergence of the imbedded planar point process
\[
N_n= \sum^n_{i=1} \delta_{({i\over n}, X^n_i)}@>D>> N\tag{\(*\)}
\]
to some Poisson (or related) point process \(N\). Here, \(X^n_i= (X_i- b_n)/a_n\) is a suitable normalization of \(X_i\). For the process \(N\) in \((*)\) which has accumulation points along a lower boundary curve, an optimal stopping problem in continuous time can be formulated. The optimal solution for the latter is given by a threshold stopping time, the threshold function being determined by a first-orer differential equation. In Section 2 it is shown that (under certain assumptions) the optimal stopping curve \(u\) solves a differential equation of the form
\[
u'(t)=- \int^\infty_{u(t)} G(t,y)\,dy,\quad 0\leq t< 1,\;u(1)= c\tag{\(**\)}
\]
for some guarantee value \(-\infty\leq c<\infty\). Here, \(G\) (called the “intensity function”) is defined explicitly via the intensity measure of \(N\). If \(c\) is finite, \((**)\) has a unique solution. If \(c=-\infty\), the authors show that the optimal stopping curve is the maximal solution of \((**)\). In Section 3 it is shown that \((**)\) can be solved in “explicit form” if \(G\) satisfies a certain “separation condition”. In Section 4 an approximation result by R. Kühne and L. Rüschendorf (loc. cit.) is extended to dependent sequences. In Section 5 optimal stopping of sequences of the form
\[
X_i= c_i Z_i+ d+i
\]
is discussed in a fairly complete form. Here, J\((Z_i)\) is an i.i.d. sequence, and the discount factors \((c_i)\) and the observation costs \((d_i)\) fulfill some criteria to ensure point process convergence.

Reviewer: Klaus Schürger (Bonn)

##### MSC:

60G40 | Stopping times; optimal stopping problems; gambling theory |

62L15 | Optimal stopping in statistics |

PDF
BibTeX
XML
Cite

\textit{A. Faller} and \textit{L. Rüschendorf}, Adv. Appl. Probab. 43, No. 4, 1086--1108 (2011; Zbl 1235.60038)

Full Text:
DOI

**OpenURL**

##### References:

[1] | Bruss, F. T. and Rogers, L. C. G. (1991). Embedding optimal selection problems in a Poisson process. Stoch. Process. Appl. 38, 267-278. · Zbl 0745.60040 |

[2] | Chow, Y. S., Robbins, H. and Siegmund, D. (1971). Great Expectations: The Theory of Optimal Stopping. Houghton Mifflin, Boston, MA. · Zbl 0233.60044 |

[3] | De Haan, L. and Verkaade, E. (1987). On extreme-value theory in the presence of a trend. J. Appl. Prob. 24, 62-76. · Zbl 0623.60035 |

[4] | Faller, A. (2009). Approximative Lösungen von Mehrfachstoppproblemen . Doctoral Thesis, University of Freiburg. · Zbl 1185.60003 |

[5] | Faller, A. and Rüschendorf, L. (2011). On approximative solutions of multistopping problems. Ann. Appl. Prob. 21 , 1965-1993. · Zbl 1251.60038 |

[6] | Gnedin, A. V. and Sakaguchi, M. (1992). On a best choice problem related to the Poisson process. In Strategies for Sequential Search and Selection in Real Time (Amherst, MA, 1990; Contemp. Math. 125 ), American Mathematical Society, Providence, RI, pp. 59-64. · Zbl 0751.60041 |

[7] | Kennedy, D. P. and Kertz, R. P. (1990). Limit theorems for threshold-stopped random variables with applications to optimal stopping. Adv. Appl. Prob. 22, 396-411. · Zbl 0704.60021 |

[8] | Kennedy, D. P. and Kertz, R. P. (1991). The asymptotic behavior of the reward sequence in the optimal stopping of i.i.d. random variables. Ann. Prob. 19, 329-341. · Zbl 0722.60036 |

[9] | Kühne, R. and Rüschendorf, L. (2000a). Approximation of optimal stopping problems. Stoch. Process. Appl. 90, 301-325. · Zbl 1047.60039 |

[10] | Kühne, R. and Rüschendorf, L. (2000b). Optimal stopping with discount and observation costs. J. Appl. Prob. 37, 64-72. · Zbl 0971.60044 |

[11] | Kühne, R. and Rüschendorf, L. (2003). Optimal stopping and cluster point processes. Statist. Decisions 21, 261-282. · Zbl 1041.60041 |

[12] | Kühne, R. and Rüschendorf, L. (2004). Approximate optimal stopping of dependent sequences. Theory Prob. Appl. 48, 465-480. · Zbl 1079.60043 |

[13] | Resnick, S. I. (1987). Extreme Values, Regular Variation, and Point Processes. Springer, New York. · Zbl 0633.60001 |

This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.