Intriguing Sum

S = n = 0 k = 0 n 2 n ( 1 ) k ( n k ) ( k + 1 ) 4 S=\sum_{n=0}^{\infty}\sum_{k=0}^n2^{-n}(-1)^k{n \choose k}(k+1)^{-4}

Write S = a π c b S=\dfrac{a\pi^c}{b} , where a , b a,b and c c are positive integers and a , b a,b are coprime. Enter a + b + c a+b+c as your answer.


The answer is 371.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

2 solutions

Mark Hennings
Feb 9, 2016

We have n = 0 k = 0 n 2 n ( 1 ) k ( n k ) x k = n = 0 2 n ( 1 x ) n = 2 1 + x , \sum_{n=0}^\infty \sum_{k=0}^n 2^{-n} (-1)^k {n \choose k}x^k \; = \; \sum_{n=0}^\infty 2^{-n}(1 - x)^n \; = \; \frac{2}{1+x} \;, and this series is absolutely convergent for 1 < x 1 -1 < x \le 1 and uniformly convergent for 0 x 1 0 \le x \le 1 . Since 0 1 x α ( ln x ) 3 d x = 6 ( α + 1 ) 4 , α > 1 , \int_0^1 x^\alpha (\ln x)^3\,dx \; = \; \frac{-6}{(\alpha+1)^4} \;, \qquad \alpha > -1 \;, we see that S = 1 6 0 1 2 1 + x ( ln x ) 3 d x = 1 3 n = 0 ( 1 ) n 0 1 x n ( ln x ) 3 d x = 2 n = 0 ( 1 ) n 1 ( n + 1 ) 4 S \; = \; -\tfrac16\int_0^1 \frac{2}{1+x} (\ln x)^3\,dx \; = \; -\tfrac13\sum_{n=0}^\infty (-1)^n \int_0^1 x^n(\ln x)^3\,dx \; = \; 2\sum_{n=0}^\infty (-1)^n \frac{1}{(n+1)^4} and so S = 2 ( n = 1 1 n 4 2 n = 1 1 ( 2 n ) 4 ) = 2 ( 1 1 8 ) ζ ( 4 ) = 7 360 π 4 , S \; = \; 2\left(\sum_{n=1}^\infty \frac{1}{n^4} - 2\sum_{n=1}^\infty \frac{1} {(2n)^4}\right) \; = \; 2\big(1 - \tfrac18)\zeta(4) \; = \; \tfrac{7}{360}\pi^4 \;, making the answer 7 + 360 + 4 = 371 7 + 360 + 4 = \boxed{371} .

Another clear and careful solution! (+1) Thanks!

Otto Bretscher - 5 years, 4 months ago

Log in to reply

Actually, I want to improve on my previous solution. Let X X be the number of tosses of a fair coin required to obtain k + 1 k+1 Heads. Then P [ X = n + 1 ] = 2 ( n + 1 ) ( n k ) n k , P[X = n+1] \; = \; 2^{-(n+1)} {n \choose k} \qquad \qquad n \ge k \;, and hence, since these probabilities must add to 1 1 : n = k 2 n ( n k ) = 2 . \sum_{n=k}^\infty 2^{-n}{n \choose k} \; = \; 2 \;. Then, reversing the order of summation (which is OK since the series are absolutely convergent) S = n = 0 k = 0 n 2 n ( 1 ) k ( n k ) 1 ( k + 1 ) 4 = k = 0 ( 1 ) k 1 ( k + 1 ) 4 n = k 2 n ( n k ) = 2 k = 0 ( 1 ) k 1 ( k + 1 ) 4 = 2 ( 1 1 8 ) ζ ( 4 ) = 7 360 π 4 \begin{array}{rcl} S & = & \displaystyle \sum_{n=0}^\infty \sum_{k=0}^n 2^{-n} (-1)^k {n \choose k} \frac{1}{(k+1)^4} \; = \; \sum_{k=0}^\infty (-1)^k \frac{1}{(k+1)^4}\sum_{n=k}^\infty 2^{-n} {n \choose k} \\ & = & \displaystyle 2\sum_{k=0}^\infty (-1)^k \frac{1}{(k+1)^4} \\ & = & 2\big(1 - \tfrac18\big)\zeta(4) \; =\; \frac{7}{360}\pi^4 \end{array}

Mark Hennings - 5 years, 4 months ago

Log in to reply

Yes, this is an interesting variant (+1)

Otto Bretscher - 5 years, 4 months ago

I wished Brilliant allows you to post your solution twice! Loved this alternative answer.

Grabs printer

Pi Han Goh - 5 years, 3 months ago

In general, n = k x n ( n k ) = x k ( 1 x ) k + 1 \sum_{n=k}^{\infty} x^{n} \dbinom{n}{k} = \dfrac{x^k}{(1-x)^{k+1}}

Ishan Singh - 4 years, 11 months ago

i did the same thing. i got stuck. i am sorry. i have a question why there are so many questions in brilliant on infinite series.

Srikanth Tupurani - 2 years, 3 months ago
Otto Bretscher
Feb 9, 2016

Using Euler's (backward) Series Transformation , with a k = 1 ( k + 1 ) 4 a_k=\frac{1}{(k+1)^4} , we find that S = 2 n = 0 ( 1 ) n ( n + 1 ) 4 S=2\sum_{n=0}^{\infty}\frac{(-1)^n}{(n+1)^4} = 2 η ( 4 ) = 2 ( 1 1 8 ) ζ ( 4 ) = 7 360 π 4 =2\eta(4)=2\left(1-\frac{1}{8}\right)\zeta(4)=\frac{7}{360}\pi^4 . The relationship between the Dirichlet Eta Function η ( s ) \eta(s) and the Zeta Function ζ ( s ) \zeta(s) was discussed here .

Can you explain how you managed to convert a double sum ( Σ Σ \Sigma\Sigma ) into a single sum? Because I don't see it anywhere in the very first link you shared.

Pi Han Goh - 5 years, 3 months ago

Log in to reply

If you combine equations (1), (4), and (5) in the link, the Euler-Knopp transformation takes the form k = 0 ( 1 ) k a k = k = 0 m = 0 k 1 2 k + 1 ( 1 ) m ( k m ) a m \sum_{k=0}^{\infty}(-1)^k a_k=\sum_{k=0}^{\infty}\sum_{m=0}^k \frac{1}{2^{k+1}}(-1)^m{k \choose m}a_m

This is all I'm using. (Euler initially introduced this kind of transformation to "accelerate convergence")

Otto Bretscher - 5 years, 3 months ago

Log in to reply

Hmmm... I've gotten another form (different from yours). I'll come back to this some other time.

Pi Han Goh - 5 years, 3 months ago

Or you can just apply equation (21) to get the answer.

I wonder how to prove that equation...

Pi Han Goh - 5 years, 3 months ago

Log in to reply

@Pi Han Goh Well, that's where my problem comes from. I was looking at this analytic continuation of ζ ( s ) \zeta(s) , and, like you, I was wondering how to prove it. The best I could think of was the Euler transform.

Otto Bretscher - 5 years, 3 months ago

Nice and unique approach. Another (similar) way is Summation by Parts (which can also be used to prove Euler's Series Transform).

Ishan Singh - 5 years, 2 months ago

Log in to reply

PROOF! PROOF! PROOF! PROOF!

Pi Han Goh - 5 years, 2 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...