MHB Can the Partial Sum of a Difficult Series be Solved with Induction?

  • Thread starter Thread starter MarkFL
  • Start date Start date
  • Tags Tags
    Partial Sum
AI Thread Summary
The discussion centers around finding the partial sum of the series S = ∑(2^k / (3^(2^k) + 1)) for k from 0 to n, specifically for n = 2006. Participants suggest calculating initial sums to observe convergence towards 1/2, which is one of the answer choices. A proposed formula for the general partial sum is S_n = 1/2 - (2^(n+1) / (3^(2^(n+1)) - 1), leading to option D as the likely correct answer. Induction is recommended as a method to prove this formula. The conversation highlights the importance of enumerating terms to derive insights into the series behavior.
MarkFL
Gold Member
MHB
Messages
13,284
Reaction score
12
Hello MHB! (Wave)

A young man in high school I know has been essentially tasked with finding the following partial sum:

$$S=\sum_{k=0}^{n}\left(\frac{2^k}{3^{2^k}+1}\right)$$

I honestly have no idea how to proceed, and I am hoping someone here can provide some insight. (Star)
 
Mathematics news on Phys.org
MarkFL said:
Hello MHB! (Wave)

A young man in high school I know has been essentially tasked with finding the following partial sum:

$$S=\sum_{k=0}^{n}\left(\frac{2^k}{3^{2^k}+1}\right)$$

I honestly have no idea how to proceed, and I am hoping someone here can provide some insight. (Star)
Hey MarkFL,

For high school I would suggest to simply write it out:
$$S_0=\frac{2^0}{3^{2^0}+1}=\frac 14 = 0.25 \\
S_1=\frac 14 + \frac{2^1}{3^{2^1}+1}=\frac 14 + \frac 2{10} = \frac 9{20} = 0.45 \\
S_2=\frac 9{20} + \frac{2^2}{3^{2^2}+1}=\frac 9{20} + \frac{4}{82} = \frac{409}{820}\approx 0.49878 \\
S_3=\frac{409}{820} + \frac{2^3}{3^{2^3}+1} = \frac{409}{820} + \frac{8}{6561} \approx 0.4999996
$$
 
I like Serena said:
Hey MarkFL,

For high school I would suggest to simply write it out:
$$S_0=\frac{2^0}{3^{2^0}+1}=\frac 14 = 0.25 \\
S_1=\frac 14 + \frac{2^1}{3^{2^1}+1}=\frac 14 + \frac 2{10} = \frac 9{20} = 0.45 \\
S_2=\frac 9{20} + \frac{2^2}{3^{2^2}+1}=\frac 9{20} + \frac{4}{82} = \frac{409}{820}\approx 0.49878 \\
S_3=\frac{409}{820} + \frac{2^3}{3^{2^3}+1} = \frac{409}{820} + \frac{8}{6561} \approx 0.4999996
$$

The actual sum he's given is:

$$S=\sum_{k=0}^{2006}\left(\frac{2^k}{3^{2^k}+1}\right)$$

And he's given several choices for the result. I was hoping to find the general partial sum in closed form, and then answer the question that way.

W|A tells me the corresponding infinite sum is 1/2, but I don't even know how to show that. (Sweating)
 
MarkFL said:
The actual sum he's given is:

$$S=\sum_{k=0}^{2006}\left(\frac{2^k}{3^{2^k}+1}\right)$$

And he's given several choices for the result. I was hoping to find the general partial sum in closed form, and then answer the question that way.

W|A tells me the corresponding infinite sum is 1/2, but I don't even know how to show that. (Sweating)

I take it that 1/2 is one of the choices?
We can show it by simply expanding the first couple of partial sums, which show that the partial sum converges pretty quickly to 1/2.
Indeed, it's not a proof, but I don't expect that to be required on high-school level. Just to be able to get the result.

Is this about a math olympiad?
I learned a long time ago that many of the problems posed can typically be resolved by just enumerating.
 
I like Serena said:
I take it that 1/2 is one of the choices?
We can show it by simply expanding the first couple of partial sums, which show that the partial sum converges pretty quickly to 1/2.
Indeed, it's not a proof, but I don't expect that to be required on high-school level. Just to be able to get the result.

Is this about a math olympiad?
I learned a long time ago that many of the problems posed can typically be resolved by just enumerating.

Here are the choices:

A.) $$\frac{1}{2}$$

B.) $$\frac{1}{2}-\frac{2^{2006}}{3^{2^{2006}}-1}$$

C.) $$\frac{1}{2}-\frac{2^{2005}}{3^{2^{2005}}-1}$$

D.) $$\frac{1}{2}-\frac{2^{2007}}{3^{2^{2007}}-1}$$

E.) None of the above.

I'm not sure where the problem came from, but it's not for a competition, just a problem given to a high school student in a gifted program.
 
I like Serena said:
For high school I would suggest to simply write it out:
$$S_0=\frac{2^0}{3^{2^0}+1}=\frac 14 = 0.25 \\
S_1=\frac 14 + \frac{2^1}{3^{2^1}+1}=\frac 14 + \frac 2{10} = \frac 9{20} = 0.45 \\
S_2=\frac 9{20} + \frac{2^2}{3^{2^2}+1}=\frac 9{20} + \frac{4}{82} = \frac{409}{820}\approx 0.49878 \\
S_3=\frac{409}{820} + \frac{2^3}{3^{2^3}+1} = \frac{409}{820} + \frac{8}{656{\color{red}2}} \approx 0.4999996
$$
Writing those as $$S_0= \frac 14 = \frac12 - \frac14 = \frac12 - \frac28 = \frac12 - \frac{2^1}{3^{2^1} - 1}, \\
S_1= \frac 9{20} = \frac12 -\frac1{20} = \frac12 - \frac4{80} = \frac12 - \frac{2^2}{3^{2^2} - 1}, \\
S_2= \frac{409}{820} = \frac12 - \frac1{820} = \frac12 - \frac8{6560} = \frac12 - \frac{2^3}{3^{2^3} - 1},$$ it begins to look as though $$ S_n = \frac12 - \frac{2^{n+1}}{3^{2^{n+1}} - 1}.$$ That suggests option D.) as the correct answer for $S_{2006}.$

If you want something more convincing, you can use induction to prove that suggested formula for $S_n$.
 
Opalg said:
Writing those as $$S_0= \frac 14 = \frac12 - \frac14 = \frac12 - \frac28 = \frac12 - \frac{2^1}{3^{2^1} - 1}, \\
S_1= \frac 9{20} = \frac12 -\frac1{20} = \frac12 - \frac4{80} = \frac12 - \frac{2^2}{3^{2^2} - 1}, \\
S_2= \frac{409}{820} = \frac12 - \frac1{820} = \frac12 - \frac8{6560} = \frac12 - \frac{2^3}{3^{2^3} - 1},$$ it begins to look as though $$ S_n = \frac12 - \frac{2^{n+1}}{3^{2^{n+1}} - 1}.$$ That suggests option D.) as the correct answer for $S_{2006}.$

If you want something more convincing, you can use induction to prove that suggested formula for $S_n$.

Awesome work, as always Chris! (Yes)

Using your post, where the base case and induction statement have been given, then the induction step would be:

$$S_{n+1}=\frac{1}{2}-\frac{2^{n+1}}{3^{2^{n+1}}-1}+\frac{2^{n+1}}{3^{2^{n+1}}+1}=\frac{1}{2}-\frac{2^{n+2}}{3^{2^{n+2}}-1}$$

And this completes the proof by induction. :)
 

Similar threads

Back
Top