Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Can someone help evaluate this limit: $$\lim_{n\rightarrow \infty} \sum_{k=0}^n\frac{k^k(n-k)^{n-k}}{k!(n-k)!}n!n^{-n-1/2}$$

share|improve this question
    
What have you tried? And if this is homework, please add the homework tag. –  Stefan Hansen Jan 29 '13 at 10:03

3 Answers 3

Stirling's approximation gives us that $$ n!=\sqrt{2\pi n}\ n^n\,e^{-n}(1+O(1/n))\tag{1} $$ As it is simply the $j=k$ term in $$ \sum_{j=0}^n\binom{n}{j}\left(\frac{k}{n}\right)^{j}\left(1-\frac{k}{n}\right)^{n-j}=1\tag{2} $$ we can deduce that $$ \binom{n}{k}\left(\frac{k}{n}\right)^{k}\left(1-\frac{k}{n}\right)^{n-k}\le1\tag{3} $$ Therefore, $$ \begin{align} &\sum_{k=0}^n\frac{k^k(n-k)^{n-k}}{k!(n-k)!}n!n^{-n-1/2}\\ &=\sum_{k< m}\binom{n}{k}\left(\frac{k}{n}\right)^{k}\left(1-\frac{k}{n}\right)^{n-k}n^{-1/2}\\ &+\sum_{k=m}^{n-m}\binom{n}{k}\left(\frac{k}{n}\right)^{k}\left(1-\frac{k}{n}\right)^{n-k}n^{-1/2}\\ &+\sum_{k>n-m}\binom{n}{k}\left(\frac{k}{n}\right)^{k}\left(1-\frac{k}{n}\right)^{n-k}n^{-1/2}\\ &=O(mn^{-1/2})+\frac1{\sqrt{2\pi}}\sum_{k=m}^{n-m}\frac1{\sqrt{k}\sqrt{n-k}}(1+O(1/m))+O(mn^{-1/2})\\ &=O(mn^{-1/2})+\frac1{\sqrt{2\pi}}\sum_{k=0}^n\frac1{\sqrt{k/n}\sqrt{1-k/n}}\frac1n(1+O(1/m))\tag{4} \end{align} $$ We can add the tails to the last sum in $(4)$ since each of the $2m$ terms is $\le n^{-1/2}$ and absorb the difference into $O(mn^{-1/2})$.

Approximating with Riemann sums, $(4)$ becomes $$ \lim_{n\to\infty}\sum_{k=0}^n\frac{k^k(n-k)^{n-k}}{k!(n-k)!}n!n^{-n-1/2} =\frac1{\sqrt{2\pi}}\int_0^1\frac{\mathrm{d}x}{\sqrt{x-x^2}}(1+O(1/m))\tag{5} $$ Since $m$ was arbitrary, we get $$ \begin{align} \lim_{n\to\infty}\sum_{k=0}^n\frac{k^k(n-k)^{n-k}}{k!(n-k)!}n!n^{-n-1/2} &=\frac1{\sqrt{2\pi}}\int_0^1\frac{\mathrm{d}x}{\sqrt{x-x^2}}\\ &=\frac1{\sqrt{2\pi}}\int_{-1}^1\frac{\mathrm{d}x}{\sqrt{1-x^2}}\\ &=\sqrt{\frac\pi2}\tag{6} \end{align} $$

share|improve this answer

Stirling's approximation shows that the limit can be rewritten $$\lim_{n\rightarrow \infty} \sum_{k=0}^n\frac{k^k(n-k)^{n-k}}{k!(n-k)!}\sqrt{2\pi}e^{-n}$$, but I don't know if that makes it easier.

share|improve this answer
    
No, it cant be written like that based on what you know, the limit of the product of two expressions only equals the product of each individual limit, if both limits exist. –  Ethan Jan 29 '13 at 11:05

Are you sure the binomial theorem works? Are you suggesting $$\sum_{k=0}^n \frac{n!}{k!(n-k)!}k^k(n-k)^{n-k} = (k+n-k)^n$$? $k$ is the dummy variable used in the sum though; it is not constant.

share|improve this answer
    
It doesn't sorry –  Ethan Jan 29 '13 at 11:01

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.