Let $x_1,x_2,\dots,x_n$ be independent identically distributed random variables uniform on $\{1,2,\dots,N\}$, and let: $Y_n:=\text{the number of different elements in } \{x_1,x_2,\dots,x_n\}$.
Let $T:=\inf\{n:Y_n=N\}$.
What is $E\left[T\right]$?
Asked
Active
Viewed 105 times
0
-
What do you mean by $\inf{n:Y_n=N}$? – Christian Chapman Feb 10 '14 at 03:42
-
1The question is asking for the expected value of the random minimum sample size $T$ needed to be observed from a discrete uniform variable on $[1, N]$ such that every value in the support is observed. Clearly, ${\rm E}[T] \ge N$. – heropup Feb 10 '14 at 03:48
-
1@enthdegree the smallest n s.t. Yn=N.i.e. the first time we have N different elements in {x1,x2,...,xn}. – user98619 Feb 10 '14 at 03:50
-
Get it now, thanks! – Christian Chapman Feb 10 '14 at 03:50
-
@heropup You are right!Thank you for the explanation – user98619 Feb 10 '14 at 03:51
-
@enthdegree It is my fault.I guess there should be some trick to solve this. – user98619 Feb 10 '14 at 03:53
1 Answers
0
This question is commonly known as the coupon collector's problem. See here: http://en.wikipedia.org/wiki/Coupon_collector%27s_problem
heropup
- 143,828