Value of summation involving combination

AI Thread Summary
The discussion revolves around solving the summation $$\sum_{k=1}^{64} {64 \choose k} 64k$$ using combinatorial techniques. Participants suggest starting with smaller values of n to identify patterns and propose using integration to simplify the summation. A key approach involves multiplying the terms by a power of s, integrating, and then differentiating to relate the results back to the original summation. Ultimately, the use of the binomial theorem leads to the conclusion that the sum evaluates to $$2^{75}$$. The conversation highlights the importance of systematic mathematical steps in deriving the final answer.
songoku
Messages
2,467
Reaction score
382

Homework Statement


Find

$$\sum_{k=1}^{64} {64 \choose k} 64k$$

Homework Equations


Not sure

The Attempt at a Solution


Please give me hint how to start doing this question
 
Physics news on Phys.org
songoku said:

Homework Statement


Find

$$\sum_{k=1}^{64} {64 \choose k} 64k$$

Homework Equations


Not sure

The Attempt at a Solution


Please give me hint how to start doing this question

Work it out for n = 1, 2, 3, 4, 5 and see whether you can spot a pattern. Then, perhaps, try induction on ##n## to prove a general formula. Finally, set ##n = 64##
 
  • Like
Likes songoku
A standard method for getting rid of an awkward factor consisting of the index variable (the k in 64k) is to make the sum a function of some unknown, s say, by multiplying the kth term by s to some power. E.g.
$$\sum_{k=1}^{64} {64 \choose k} 64k s^k$$
What would happen if you tried to integrate that wrt s? Ok, so try a slightly different power.
 
  • Like
Likes songoku
songoku said:

Homework Statement


Find

$$\sum_{k=1}^{64} {64 \choose k} 64k$$

Homework Equations


Not sure

The Attempt at a Solution


Please give me hint how to start doing this question

Here is a little hint for a non-calculus solution, using n = 4 instead of n = 64. Note that for ##k =0## we have ##0 {4 \choose 0} = 0##, so we can start the summation from ##k = 0## instead of ##k = 1##. Let ##S_4 = \sum_{k=0}^4 k {4 \choose k}##. We have
$$S_4 = 0 {4 \choose 0} + 1 {4 \choose 1} + 2 {4 \choose 2} + 3 {4 \choose 3} + 4 {4 \choose 4}$$.
However, we have ##{n \choose k} = {n \choose n-k}## for all ##k##, so we can write
$$S_4 = 0 {4 \choose 4} + 1 {4 \choose 3} + 2 {4 \choose 2} + 3 {4 \choose 1} + 4 {4 \choose 0} $$.
Now add the two forms together, collecting coefficients of each ##{4 \choose j}##. That will give you
$$2S_4 = (0+4) {4 \choose 0} + (1+3) {4 \choose 1} + \cdots + (4 + 0) {4 \choose 4} = 4 \sum_{k=0}^4 {4 \choose k}.$$
Do you recognize that last summation?
 
Last edited by a moderator:
  • Like
Likes songoku
PeroK said:
Work it out for n = 1, 2, 3, 4, 5 and see whether you can spot a pattern. Then, perhaps, try induction on ##n## to prove a general formula. Finally, set ##n = 64##

Maybe the pattern you are referring is the same as Ray Vickson?

Ray Vickson said:
Here is a little hint for a non-calculus solution, using n = 4 instead of n = 64. Note that for ##k =0## we have ##0 {4 \choose 0} = 0##, so we can start the summation from ##k = 0## instead of ##k = 1##. Let ##S_4 = \sum_{k=0}^4 k {4 \choose k}##. We have
$$S_4 = 0 {4 \choose 0} + 1 {4 \choose 1} + 2 {4 \choose 2} + 3 {4 \choose 3} + 4 {4 \choose 4}$$.
However, we have ##{n \choose k} = {n \choose n-k}## for all ##k##, so we can write
$$S_4 = 0 {4 \choose 4} + 1 {4 \choose 3} + 2 {4 \choose 2} + 3 {4 \choose 1} + 4 {4 \choose 0} $$.
Now add the two forms together, collecting coefficients of each ##{4 \choose j}##. That will give you
$$2S_4 = (0+4) {4 \choose 0} + (1+3) {4 \choose 1} + \cdots + (4 + 0) {4 \choose 4} = 4 \sum_{k=0}^4 {4 \choose k}.$$
Do you recognize that last summation?

I get your hint. Thanks

haruspex said:
A standard method for getting rid of an awkward factor consisting of the index variable (the k in 64k) is to make the sum a function of some unknown, s say, by multiplying the kth term by s to some power. E.g.
$$\sum_{k=1}^{64} {64 \choose k} 64k s^k$$
What would happen if you tried to integrate that wrt s? Ok, so try a slightly different power.

I am not sure I get your hint but I am interested to know how to solve the question using integration.

I try to use sk-1 and after integration with respect to s, I get 64Ck . 64 . sk
I am guessing I need to use upper and lower limit for the integration (and maybe it is 1 and 0). But I think I am missing something that relates the summation and the integration.

Maybe I need to find a form similar to:

$$\sum_{k=1}^{64} {64 \choose k} 64k = \int_{a}^{b} {64 \choose k} 64k s^{k-1} ds $$

but I am not sure how
 
songoku said:
I try to use sk-1 and after integration with respect to s, I get 64Ck . 64 . sk
Yes, that's the right step.
Do you not recognise that sum? (Binomial Theorem.)
 
haruspex said:
Yes, that's the right step.
Do you not recognise that sum? (Binomial Theorem.)

I recognise that sum, sum of coefficient of pascal triangle which will lead to 2n

What I am confused about is how to write down the systematic mathematical step and how it leads to a proper presentation of answer.

I start by denoting 64Ck . 64k as Uk.

Uk = 64Ck . 64k

Then multiplying the term with sk-1, it becomes:

Uk = 64Ck . 64k . sk-1

Then integrating both sides, I get:

$$\int_{0}^{1} U_{k} ~ ds = \int_{0}^{1} {64 \choose k} 64k s^{k-1} ds $$

$$\int_{0}^{1} U_{k} ~ ds = {64 \choose k} 64$$

Then, how I relate the result of integration to the sum that the question asked? I mean, my result does not have sigma sign on the RHS so how can I make it becomes like:

$$ {64 \choose k} 64 = \sum_{k=0}^{k=64} {64 \choose k} 64 $$
 
songoku said:
integrating both sides, I get:
Keep it as an indefinite integral. You will need to differentiate wrt s later.
 
  • Like
Likes songoku
haruspex said:
Keep it as an indefinite integral. You will need to differentiate wrt s later.

So:
$$\int U_{k} ~ ds = \int {64 \choose k} 64k s^{k-1} ds $$

$$ = {64 \choose k} 64 s^{k} $$

Then I differentiate it with respect to s? Wouldn't it become the form before integration?
 
  • #10
songoku said:
Then I differentiate it with respect to s?
Not yet. Sum it first.
The sequence is: multiply by sk-1, integrate, perform the sum, differentiate wrt s, plug in s=1.
 
  • Like
Likes songoku
  • #11
haruspex said:
Not yet. Sum it first.
The sequence is: multiply by sk-1, integrate, perform the sum, differentiate wrt s, plug in s=1.
So:
$$\int U_{k} ~ ds = {64 \choose k} 64 s^{k} $$
$$\sum_{k=0}^{k=64} \int U_{k} ~ ds =\sum_{k=0}^{k=64} {64 \choose k} 64 s^{k} $$
$$=64 ({64 \choose 0} s^{0} + {64 \choose 1} s^{1} + ... + {64 \choose 64} s^{64}) $$

Differentiate both sides with respect to s:
$$\sum_{k=0}^{k=64} U_{k} = {64 \choose 1} + 2 {64 \choose 2} s + 3 {64 \choose 3} s^{2} + ... + 64 {64 \choose 64} s^{63} $$

Put s = 1:
$$\sum_{k=0}^{k=64} U_{k} = {64 \choose 1} + 2 {64 \choose 2} + 3 {64 \choose 3} + ... + 64 {64 \choose 64} $$

Am I doing it correctly?
 
  • #12
songoku said:
Am I doing it correctly?
No, you have not performed the sum. You have left it as a Σ. It sums easily into a closed form.
You are familiar with the binomial theorem, yes?
 
  • #13
haruspex said:
No, you have not performed the sum. You have left it as a Σ. It sums easily into a closed form.
You are familiar with the binomial theorem, yes?

I have learned about binomial theorem but I am trying to figure out how to use it for this question.

Let me try again:
$$\int U_{k} ~ ds = {64 \choose k} 64 s^{k} $$
$$\sum_{k=0}^{k=64} \int U_{k} ~ ds =\sum_{k=0}^{k=64} {64 \choose k} 64 s^{k} $$
$$=64 (1+s)^{64}$$

Differentiate both sides with respect to s:
$$\sum_{k=0}^{k=64} U_{k} = 64^{2} (1+s)^{63}$$

Put s = 1:
$$\sum_{k=0}^{k=64} U_{k} = 64^{2} (2)^{63}$$
$$=2^{75}$$

Is this how I suppose to do it?
 
  • #14
songoku said:
I have learned about binomial theorem but I am trying to figure out how to use it for this question.

Let me try again:
$$\int U_{k} ~ ds = {64 \choose k} 64 s^{k} $$
$$\sum_{k=0}^{k=64} \int U_{k} ~ ds =\sum_{k=0}^{k=64} {64 \choose k} 64 s^{k} $$
$$=64 (1+s)^{64}$$

Differentiate both sides with respect to s:
$$\sum_{k=0}^{k=64} U_{k} = 64^{2} (1+s)^{63}$$

Put s = 1:
$$\sum_{k=0}^{k=64} U_{k} = 64^{2} (2)^{63}$$
$$=2^{75}$$

Is this how I suppose to do it?
You got it.
 
  • #15
Thanks a lot for the help (Perok, Ray Vickson, haruspex)
 
Back
Top