Prove that [itex]\sum_n \frac{1}{n^2} < 2[/itex]

  • Thread starter Thread starter Portuga
  • Start date Start date
Click For Summary

Homework Help Overview

The problem involves proving that the series \( \sum_{n=1}^{\infty} \frac{1}{n^2} < 2 \). This falls under the subject area of series convergence and mathematical analysis.

Discussion Character

  • Exploratory, Assumption checking, Mathematical reasoning

Approaches and Questions Raised

  • Participants discuss various attempts to establish upper bounds for the series, including comparisons with other series and functions. Some express uncertainty about the effectiveness of their approaches, while others suggest using induction or comparing the series to a continuous function.

Discussion Status

The discussion is ongoing, with participants exploring different strategies and questioning the validity of their assumptions. Some have proposed using graphical methods or telescoping series, while others are considering the implications of their mathematical inequalities.

Contextual Notes

There are indications of confusion regarding the setup of the problem and the relevant equations. Participants are also navigating the constraints of homework rules that may limit the methods they can use.

Portuga
Messages
56
Reaction score
6

Homework Statement



Prove that [tex]1+\frac{1}{2^{2}}+\frac{1}{3^{2}}+\ldots+\frac{1}{n^{2}}<2.[/tex]

Homework Equations



[tex]1 = 2^0, 2 = 2^1, 3 = 2^2 - 1, 4 = 2^2, \ldots,[/tex]

and

[tex]2^{n}<2^{n}+1<\ldots<2^{n+1}-1[/tex]

The Attempt at a Solution


As [tex]2^{n}<2^{n}+1<\ldots<2^{n+1}-1,[/tex] it's true that
[tex] \frac{1}{2^{n}}>\frac{1}{2^{n}+1}>\ldots\frac{1}{2^{n+1}-1}.[/tex]
So, being
[tex] S_{n}=\left(\frac{1}{2^{n}}\right)\left(\frac{1}{2^{n}}\right)+\left(\frac{1}{2^{n}+1}\right)\left(\frac{1}{2^{n}+1}\right)+\ldots+\left(\frac{1}{2^{n+1}-1}\right)\left(\frac{1}{2^{n+1}-1}\right),[/tex]
follows from previous inequalities that

[tex]S_{n} < \left(\frac{1}{2^{n}}\right)\left(\frac{1}{2^{n}}\right)+\ldots+\left(\frac{1}{2^{n}}\right)\left(\frac{1}{2^{n}}\right)[/tex]

[tex]=\frac{n}{2^{2n}}.[/tex]

This is a dead-end to me, because I can't compare it with 2 in an elegant way. I made the graphic of this function, [tex]f(x) = x/2^{2x},[/tex] and I verified it's less than 2 for all x, but I am pretty sure there is a better way to do this proof.
 
Physics news on Phys.org
Portuga said:

Homework Statement



Prove that [tex]1+\frac{1}{2^{2}}+\frac{1}{3^{2}}+\ldots+\frac{1}{n^{2}}<2.[/tex]

Homework Equations



[tex]1 = 2^0, 2 = 2^1, 3 = 2^2 - 1, 4 = 2^2, \ldots,[/tex]

and

[tex]2^{n}<2^{n}+1<\ldots<2^{n+1}-1[/tex]

The Attempt at a Solution


As [tex]2^{n}<2^{n}+1<\ldots<2^{n+1}-1,[/tex] it's true that
[tex] \frac{1}{2^{n}}>\frac{1}{2^{n}+1}>\ldots\frac{1}{2^{n+1}-1}.[/tex]
So, being
[tex] S_{n}=\left(\frac{1}{2^{n}}\right)\left(\frac{1}{2^{n}}\right)+\left(\frac{1}{2^{n}+1}\right)\left(\frac{1}{2^{n}+1}\right)+\ldots+\left(\frac{1}{2^{n+1}-1}\right)\left(\frac{1}{2^{n+1}-1}\right),[/tex]
follows from previous inequalities that

[tex]S_{n} < \left(\frac{1}{2^{n}}\right)\left(\frac{1}{2^{n}}\right)+\ldots+\left(\frac{1}{2^{n}}\right)\left(\frac{1}{2^{n}}\right)[/tex]

[tex]=\frac{n}{2^{2n}}.[/tex]

This is a dead-end to me, because I can't compare it with 2 in an elegant way. I made the graphic of this function, [tex]f(x) = x/2^{2x},[/tex] and I verified it's less than 2 for all x, but I am pretty sure there is a better way to do this proof.

Are you familiar with induction?
 
  • Like
Likes   Reactions: Portuga
Assume that it is greater than or equal 2 and then show that can't be true

[tex]\sum \frac{1}{n^2}=\left(1+\frac{1}{2^2}+\frac{1}{3^2}+...+\frac{1}{n^2}\right)\ge 2<br /> \\<br /> \frac{1}{2^2}+\frac{1}{3^2}+...+\frac{1}{n^2}\ge 1<br /> \\<br /> \frac{2^2+3^2+4^2+...+n^2}{2^23^24^2...n^2}\ge 1<br /> \\[/tex]

which you can show isn't true using a few different arguments
 
RedDelicious said:
Assume that it is greater than or equal 2 and then show that can't be true

[tex]\sum \frac{1}{n^2}=\left(1+\frac{1}{2^2}+\frac{1}{3^2}+...+\frac{1}{n^2}\right)\ge 2<br /> \\<br /> \frac{1}{2^2}+\frac{1}{3^2}+...+\frac{1}{n^2}\ge 1<br /> \\<br /> \frac{2^2+3^2+4^2+...+n^2}{2^23^24^2...n^2}\ge 1<br /> \\[/tex]

which you can show isn't true using a few different arguments

[tex]\frac{1}{2^2}+\frac{1}{3^2}+...+\frac{1}{n^2} = \frac{2^2+3^2+4^2+...+n^2}{2^23^24^2...n^2} \:\Longleftarrow\:\text{FALSE!}[/tex]
 
  • Like
Likes   Reactions: RedDelicious and Portuga
Math_QED said:
Are you familiar with induction?
Yes, I will try this strategy.
 
Ray Vickson said:
[tex]\frac{1}{2^2}+\frac{1}{3^2}+...+\frac{1}{n^2} = \frac{2^2+3^2+4^2+...+n^2}{2^23^24^2...n^2} \:\Longleftarrow\:\text{FALSE!}[/tex]
Thank you very much!
 
Your Sn only sums the 1/r2 for r from n to 2n. You need the sum all the way from 1.

With regard to induction,you would need to set up your inductive hypothesis cleverly. Starting with "suppose the sum of the first N terms < 2" will get you nowhere. You'd need to find some other series summing to at most 2 and show that it acts as an upper bound for the given sequence.

A useful strategy sometimes is to compare the series with the corresponding continuous function.
 
Last edited:
Thanks. I will find one.
 
I see an issue with the first of the relevant equations, based on the goal of this problem, the first relevant equation should be:

$$ 1 = 2^0, \ 2 = 2^1, 3 = 2^1+1, \ 4 = 2^2, 5 = 2^2+1, 6 = 2^2+2, 7 = 2^2+3, 8 = 2^3, \ \ldots $$

I don't think mathematical induction will be part of the solution. I'll wait for a reply before providing any more hints.
 
Last edited:
  • #10
haruspex said:
A useful strategy sometimes is to compare the series with the corresponding continuous function.
Yes, it is an excellent hint!
Compare the sum to the definite integral of 1/x2 from 1 to n.
upload_2016-7-17_7-7-47.png
 
  • Like
Likes   Reactions: Delta2
  • #11
ehild said:
Yes, it is an excellent hint!
Compare the sum to the definite integral of 1/x2 from 1 to n.
View attachment 103343
I was avoiding being too specific at this stage.
 
  • #12
haruspex said:
I was avoiding being too specific at this stage.
Your hint "compare the series with the corresponding continuous function" was a bit misleading. I think you meant the series Σ1/n2 compared to the integral of the corresponding continuous function. With the figure, I wanted to make your hint clear .
 
  • #13
ehild said:
Your hint "compare the series with the corresponding continuous function" was a bit misleading. I think you meant the series Σ1/n2 compared to the integral of the corresponding continuous function. With the figure, I wanted to make your hint clear .
Again, I was leaving it to the OP to make the analogy between the sum of the series and the integral of the function.
 
  • #14
haruspex said:
Again, I was leaving it to the OP to make the analogy between the sum of the series and the integral of the function.
Series is the sum of a sequence. https://en.wikipedia.org/wiki/Series_(mathematics). I think you meant the sum of the sequence 1/n2.
 
  • #16
Combining the attempt in post #1 and the relevant equation in post #9:

$$1+\frac{1}{2^{2}}+\frac{1}{3^{2}}+\frac{1}{4^{2}}+\ldots < 1+\left(\frac{1}{2^{2}}+\frac{1}{2^{2}}\right)+\left(\frac{1}{4^{2}}+\frac{1}{4^{2}}+\frac{1}{4^{2}}+\frac{1}{4^{2}}\right)+ \ldots$$

A finite version of the right hand series is equal to:

$$1+\frac{1}{2}+\frac{1}{4}+\frac{1}{8}+\ldots+\frac{1}{N}$$

Multiply by 2

$$2+\frac{2}{2}+\frac{2}{4}+\frac{2}{4}+\ldots+\frac{2}{N} = 3+\frac{1}{2}+\frac{1}{4}+\ldots+\frac{2}{N}$$

Then subtract 2 x series - 1 x series to eliminate the inner terms:

$$\left(3+\frac{1}{2}+\frac{1}{4}+\ldots+\frac{2}{N}\right) - \left(1+\frac{1}{2}+\frac{1}{4}+\ldots+\frac{1}{N}\right) = 2 - \frac{1}{N}$$

Since

$$1+\frac{1}{2^{2}}+\frac{1}{3^{2}}+\frac{1}{4^{2}}+\ldots < 1+\left(\frac{1}{2^{2}}+\frac{1}{2^{2}}\right)+\left(\frac{1}{4^{2}}+\frac{1}{4^{2}}+\frac{1}{4^{2}}+\frac{1}{4^{2}}\right)+ \ldots$$

Then even an infinite series of the left hand side is < 2.
 
Last edited:
  • #17
Portuga said:

Homework Statement



Prove that [tex]1+\frac{1}{2^{2}}+\frac{1}{3^{2}}+\ldots+\frac{1}{n^{2}}<2.[/tex]

Homework Equations



[tex]1 = 2^0, 2 = 2^1, 3 = 2^2 - 1, 4 = 2^2, \ldots,[/tex]

and

[tex]2^{n}<2^{n}+1<\ldots<2^{n+1}-1[/tex]

The Attempt at a Solution


As [tex]2^{n}<2^{n}+1<\ldots<2^{n+1}-1,[/tex] it's true that
[tex] \frac{1}{2^{n}}>\frac{1}{2^{n}+1}>\ldots\frac{1}{2^{n+1}-1}.[/tex]
So, being
[tex] S_{n}=\left(\frac{1}{2^{n}}\right)\left(\frac{1}{2^{n}}\right)+\left(\frac{1}{2^{n}+1}\right)\left(\frac{1}{2^{n}+1}\right)+\ldots+\left(\frac{1}{2^{n+1}-1}\right)\left(\frac{1}{2^{n+1}-1}\right),[/tex]
follows from previous inequalities that

[tex]S_{n} < \left(\frac{1}{2^{n}}\right)\left(\frac{1}{2^{n}}\right)+\ldots+\left(\frac{1}{2^{n}}\right)\left(\frac{1}{2^{n}}\right)[/tex]

[tex]=\frac{n}{2^{2n}}.[/tex]

This is a dead-end to me, because I can't compare it with 2 in an elegant way. I made the graphic of this function, [tex]f(x) = x/2^{2x},[/tex] and I verified it's less than 2 for all x, but I am pretty sure there is a better way to do this proof.

You could also use the fact that for ##n \geq 2## we have
[tex]\frac{1}{n^2} < \frac{1}{n(n-1)} = \frac{1}{n-1} - \frac{1}{n}[/tex]
The series
$$\sum_{n=2}^N \left( \frac{1}{n-1} - \frac{1}{n} \right) $$
is easy to deal with (being a so-called "telescoping" series).
 

Similar threads

Replies
5
Views
2K
  • · Replies 17 ·
Replies
17
Views
2K
Replies
4
Views
3K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 8 ·
Replies
8
Views
1K
  • · Replies 18 ·
Replies
18
Views
3K
Replies
12
Views
4K
  • · Replies 8 ·
Replies
8
Views
2K
  • · Replies 21 ·
Replies
21
Views
2K
  • · Replies 5 ·
Replies
5
Views
1K