返回题库

和超阈 III

Sum Exceedance III

专题
Probability / 概率
难度
L6

题目详情

概率题:和超阈 III。

英文原题

Define X1,X2,⋯∼Unif(0,1)X_1, X_2, \dots \sim \text{Unif}(0,1) IID

N1=min{n:X1+⋯+Xn>1}N_1 = \text{min}\{n : X_1 + \dots + X_n > 1\}, and

Sn=X1+⋯+XnS_n = X_1 + \dots + X_n. Compute E[SN1]\mathbb{E}[S_{N_1}]. The answer will be in the form cece for a rational number cc. Find cc.

解析

设当前部分和为 x∈[0,1)x\in[0,1) 时,最终和的期望为 f(x)f(x)。

下一次加上 Y∼Unif(0,1)Y\sim\mathrm{Unif}(0,1):若 x+Y>1x+Y>1 则结束并取 x+Yx+Y;否则进入状态 x+Yx+Y。于是

f(x)=∫01−xf(x+y) dy+∫1−x1(x+y) dy.f(x)=\int_0^{1-x} f(x+y)\,dy+\int_{1-x}^1 (x+y)\,dy.

对该积分方程求解可得

f(0)=e2.f(0)=\frac{e}{2}.

因此

E[SN1]=e2=(12)e,\mathbb{E}[S_{N_1}]=\frac{e}{2}=\left(\frac12\right)e,

故 c=12c=\boxed{\frac12}。


英文解析

If the current section and are x∈[0,1)x\in[0,1), the final and are expected to be f(x)f(x).

Next add Y∼Unif(0,1)Y\sim\mathrm{Unif}(0,1): end with x+Yx+Yif x+Y>1x+Y>1; otherwise enter status x+Yx+Y.

f(x)=∫01−xf(x+y) dy+∫1−x1(x+y) dy.f(x)=\int_0^{1-x} f(x+y)\,dy+\int_{1-x}^1 (x+y)\,dy.

The integral equation can be solved

f(0)=e2.f(0)=\frac{e}{2}.

Therefore,

E[SN1]=e2=(12)e,\mathbb{E}[S_{N_1}]=\frac{e}{2}=\left(\frac12\right)e,

So c=12c=\boxed{\frac12}.