arXivDaily arXiv每日学术速递 周一至周五更新
arXiv math/0503589math.PR
收藏

[0,1]值独立同分布的差分不等式有观测值代价的随机变量

Difference prophet inequalities for [0,1]-valued i.i.d. random variables with cost for observations

Holger Kosters

摘要:

设X_1,X_2,.是[0,1]-值i.i.d.的序列。随机变量,设c \geq 0为每个观测的抽样成本,设Y_i=X_i-ic,i= 1,2,.对于n= 1,2,.,设M(Y_1,...,Y_n)=E(max_{1\leq i\leq n} Y_i)和V(Y_1,.,Y_n)= sup_{τ\in C^n} E(Y_τ),其中C_n表示Y_1,.,Y_n。对于差M(Y_1,.,Y_n)-V(Y_1,.,Y_n)的一个新的解。

英文摘要:

Let X_1,X_2,... be a sequence of [0,1]-valued i.i.d. random variables, let c\geq 0 be a sampling cost for each observation and let Y_i=X_i-ic, i=1,2,.... For n=1,2,..., let M(Y_1,...,Y_n)=E(max_{1\leq i\leq n}Y_i) and V(Y_1,...,Y_n)=sup_{τ\in C^n}E(Y_τ), where C^n denotes the set of all stopping rules for Y_1,...,Y_n. Sharp upper bounds for the difference M(Y_1,...,Y_n)-V(Y_1,...,Y_n) are given under various restrictions on c and n.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑