Inspired by Aareyan Manzoor, again

Algebra Level 4

S = k = 1 2016 x k x k + 1 S=\large \sum_{k=1}^{2016}x_kx_{k+1}

Find the maximum of S S if k = 1 2016 x k 2 = 2016 \displaystyle \sum_{k=1}^{2016}x_k^2=2016 and x 2017 = x 1 x_{2017}=x_1 .


Inspiration .


The answer is 2016.

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

Rishabh Jain
Jan 7, 2016

Applying Cauchy Shwarz inequality, \color{#20A900}{\text{Applying Cauchy Shwarz inequality,}} S = k = 1 2016 x k x k + 1 S=\sum_{k=1}^{2016}x_kx_{k+1} k = 1 2016 ( x k ) 2 k = 1 ( x k + 1 2016 ) 2 \leq \sqrt{ \sum_{k=1}^{2016}(x_k)^2 \sum_{k=1}(x_{k+1}^{2016})^2} = 2016 2 = 2016 ( S i n c e k = 1 2016 ( x k ) 2 = k = 1 2016 ( x k + 1 ) 2 = 2016 ) =\sqrt{{2016}^2}=2016 \space (Since \sum_{k=1}^{2016}(x_k)^2=\sum_{k=1}^{2016}(x_{k+1})^2=2016) Equality occurs when x 1 x 2 = x 2 x 3 = . . . . . = x 2016 x 2017 {\text{Equality occurs when}\color{#69047E}{ \frac{x_1}{x_2}=\frac{x_2}{x_3}=.....= \frac {x_{2016}}{x_{2017}}}} Or x 1 = x 2 = . . . . = x 2016 = 1 x_1=x_2=....=x_{2016}=1

same here(+1). this turns out to be easier than the original problem.

@Otto Bretscher , i dont remember "inspiring" any of you previous problems.

Aareyan Manzoor - 5 years, 5 months ago

Log in to reply

It becomes harder if you omit the "link" x 2016 x 1 x_{2016}x_1 . Try it!

I have now written a few problems on this topic @Aareyan Manzoor Enjoy!

Otto Bretscher - 5 years, 5 months ago
Otto Bretscher
Jan 7, 2016

A slight variant for those who dislike Cauchy-Schwarz (all sums run from 1 to 2016): 0 ( x k x k + 1 ) 2 = x k 2 2 x k x k + 1 + x k + 1 2 = 2 x k 2 2 x k x k + 1 0\leq \sum(x_k-x_{k+1})^2=\sum x_k^2-2\sum x_kx_{k+1}+\sum x_{k+1}^2=2\sum x_k^2-2\sum x_kx_{k+1} so x k x k + 1 x k 2 = 2016 \sum x_kx_{k+1}\leq \sum x_k^2=\boxed{2016}

You still have to show that the this value is achievable.

Pi Han Goh - 5 years, 5 months ago

Log in to reply

It seems self-evident from the inequality 0 ( x k x k + 1 ) 2 0\leq \sum(x_k-x_{k+1})^2

;)

Otto Bretscher - 5 years, 5 months ago

Log in to reply

It's not that self-evident.

Pi Han Goh - 5 years, 5 months ago

Log in to reply

@Pi Han Goh What about this one , Comrade Pi Han Goh?

Otto Bretscher - 5 years, 5 months ago

Log in to reply

@Otto Bretscher Must I use quadratic forms to get the answer too?

Pi Han Goh - 5 years, 5 months ago

Log in to reply

@Pi Han Goh Do it any way you can, Comrade... "Das Wesen der Mathematik liegt in ihrer Freiheit" (Cantor)

Otto Bretscher - 5 years, 5 months ago

Log in to reply

@Otto Bretscher What I mean to ask is "What is the solution you had in mind?" Classical inequalities? Highly doubt it. Lagrange Multipliers? don't think so.

Pi Han Goh - 5 years, 5 months ago

Log in to reply

@Pi Han Goh Classical inequalities are too weak for a problem like this , I believe (maybe somebody will prove me wrong). As you know, quadratic forms and Lagrange multipliers are equivalent in this case; either way you have to think about a matrix. This may require a bit of research: Find out what is known about the eigenvalues of a matrix "of this form". You can do it, Comrade!

Otto Bretscher - 5 years, 5 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...