Induction with binomial coefficient

In summary, "Induction with binomial coefficient" refers to a mathematical proof technique that utilizes the properties of binomial coefficients to establish the truth of a given statement for all natural numbers. The process typically involves two main steps: the base case, where the statement is verified for an initial value, and the inductive step, where it is shown that if the statement holds for an arbitrary case, it also holds for the next case. This approach highlights the recursive nature of binomial coefficients, often expressed through identities like Pascal's triangle, and demonstrates their usefulness in combinatorial proofs and problem-solving.
  • #1
Lambda96
223
75
Homework Statement
##\sum\limits_{k=0}^{n} \frac{(-1)^k}{k+1} \binom{n}{k}=\frac{1}{n+1}##
Relevant Equations
none
Hi,

I'm having problems with the proof for the induction of the following problem: ##\sum\limits_{k=0}^{n} \frac{(-1)^k}{k+1} \binom{n}{k}=\frac{1}{n+1}## with ##n \in \mathbb{N}##

I proceeded as follows:

$$\sum\limits_{k=0}^{n+1} \frac{(-1)^k}{k+1} \binom{n+1}{k}=\frac{1}{n+2}$$
$$\sum\limits_{k=0}^{n+1} \frac{(-1)^k}{k+1} \biggl(\binom{n}{k} +\binom{n}{k-1} )=\frac{1}{n+2}$$
$$\sum\limits_{k=0}^{n} \frac{(-1)^k}{k+1} \binom{n}{k} +\sum\limits_{k=0}^{n+1} \frac{(-1)^k}{k+1}\binom{n}{k-1} =\frac{1}{n+2}$$
$$ \frac{1}{n+1} + \sum\limits_{k=0}^{n+1} \frac{(-1)^k}{k+1} \binom{n}{k-1}=\frac{1}{n+2}$$

Unfortunately, I can't get any further now because I don't know how to solve the term ##\sum\limits_{k=0}^{n+1} \frac{(-1)^k}{k+1} \binom{n}{k-1}## and how to simplify it further. According to mathematica, the term would be ##-\frac{1}{n^2+3n+2}## and if I calculate ##\frac{1}{n+1}-\frac{1}{n^2+3n+2}=\frac{1}{n+2}##, I get the required result.
 
Physics news on Phys.org
  • #2
What about a direct proof? I'm on my phone, so these expressions are hard to post!
 
  • Like
Likes Lambda96
  • #3
Start with this

Lambda96 said:
##\sum\limits_{k=0}^{n} \frac{(-1)^k}{k+1} \binom{n}{k}##
Take out a factor of ##\frac1 {n+1}## and then the resulting sum must be ##1##, which I think is easy to prove.

You can then use this to get the other identity above that you needed for induction!
 
  • Like
Likes Lambda96
  • #4
A general note. You shouldn't use the equality sign if you still have to prove equality! So your proofwriting should be
\begin{align*}
\sum_{k=0}^{n+1}\dfrac{(-1)^k}{k+1}\binom{n+1}{k} &=1+\dfrac{(-1)^{n+1}}{n+2}+\sum_{k=1}^{n}\dfrac{(-1)^k}{k+1}\binom{n+1}{k}\\
&=1+\dfrac{(-1)^{n+1}}{n+2}+\sum_{k=1}^{n}\dfrac{(-1)^k}{k+1}\left[\binom{n}{k}+\binom{n}{k-1}\right]\\
&=1+\sum_{k=1}^{n}\dfrac{(-1)^k}{k+1}\binom{n}{k}-\dfrac{(-1)^{n}}{n+2}+\sum_{k=1}^{n}\dfrac{(-1)^k}{k+1}\binom{n}{k-1}\\
&=\sum_{k=0}^{n}\dfrac{(-1)^k}{k+1}\binom{n}{k}-\dfrac{(-1)^{n}}{n+2}-\sum_{j=0}^{n-1}\dfrac{(-1)^j}{j+2}\binom{n}{j}\\
&=\dfrac{1}{n+1}-\dfrac{(-1)^{n}}{n+2}-\sum_{j=0}^{n-1}\dfrac{(-1)^j}{j+2}\binom{n}{j}\\
&=\dfrac{1}{n+1}-\sum_{j=0}^{n}\dfrac{(-1)^j}{j+2}\binom{n}{j}\\
&=\ldots\\
&\phantom{=}\vdots\\
&=\dfrac{1}{n+2}
\end{align*}
where ##1/(n+2)## only occurs when you actually arrived there.

I only wrote with a grain of salt what you already had, so it'll be no help. I got stuck in the same position.
A direct proof as @PeroK has suggested is definitely worth a try.
 
Last edited:
  • Informative
  • Like
Likes Lambda96 and berkeman
  • #5
Integrate from 0 to 1 the identity

##(1-x)^n=\sum\limits_{k=0}^{n} \binom{n}{k} (-1)^kx^k##
 
  • Like
Likes Lambda96 and PeroK
  • #6
Thank you PeroK, fresh_42 and martinbn for your help.

I wanted to use the tip from PeroK, with the factoring out of ##\frac{1}{n+1}## and got the following:

$$\sum\limits_{k=0}^{n}\frac{(-1)^k}{k+1} \binom{n}{k}=\sum\limits_{k=0}^{n}\frac{(-1)^k}{k+1} \frac{k+1}{n+1} \binom{n+1}{k+1}$$
$$\sum\limits_{k=0}^{n}\frac{(-1)^k}{k+1} \binom{n}{k} = \frac{1}{n+1} \sum\limits_{k=0}^{n}(-1)^k \binom{n+1}{k+1}$$

Unfortunately, I now have problems showing that the term ##\sum\limits_{k=0}^{n}(-1)^k \binom{n+1}{k+1}## is equal to 1.

@martinbn Did you mean the following ##\int_{0}^{1} (1-x)^n \,dx##?
 
  • #7
\begin{align*}
0&=(1-1)^{n+1}=\sum_{j=0}^{n+1} \binom{n+1}{j}1^{n+1-j}(-1)^j\\
&=1+\sum_{j=1}^{n+1} \binom{n+1}{j}1^{n+1-j}(-1)^j=1+\sum_{k=0}^n \binom{n+1}{k+1}(-1)^{k+1}
\end{align*}

That is what I wanted to show you in my first reply.
  • Changing the index variable by substitutions of the kind ##k \leftrightarrows j\pm 1## is often necessary when dealing with such sums.
  • The boundaries at the bottom and at the top of sums often require a specific treatment if we substitute the index since they are trivially right or wrong, or carry negative indices after the substitution which we do not want.
 
  • Like
Likes Lambda96 and PeroK
  • #8
Lambda96 said:
Unfortunately, I now have problems showing that the term ##\sum\limits_{k=0}^{n}(-1)^k \binom{n+1}{k+1}## is equal to 1.
i thought you'd see that is closely related to ##(1 -1)^{n+1}##
 
  • Like
Likes Lambda96
  • #9
Lambda96 said:
@martinbn Did you mean the following ##\int_{0}^{1} (1-x)^n \,dx##?
Let's put it the other way around: what do you get if you integrate the equation?
 
  • Like
Likes Lambda96
  • #10
Interesting. Maybe it would help to find the general form for ##\frac{1}{k}-\frac{1}{k+1}##, for context.

Edit ## n^2+3n+2=(n+1)(n+2)##
 
  • Like
Likes Lambda96
  • #11
Lambda96 said:
Homework Statement: ##\sum\limits_{k=0}^{n} \frac{(-1)^k}{k+1} \binom{n}{k}=\frac{1}{n+1}##
Here's a generalisation that you may want to try to prove. For ##m \ge 1##:
$$\sum\limits_{k=0}^{n}\frac{(-1)^k}{k+m}\binom n k = \frac{(m -1)!}{(n+1)\dots(n+m)}$$Your identity is the case for ##m =1## and the next one is:
$$\sum\limits_{k=0}^{n}\frac{(-1)^k}{k+2}\binom n k = \frac 1 {(n+1)(n+2)}$$
 
  • Like
Likes Lambda96 and fresh_42
  • #12
Thank you PeroK, fresh_42 and WWGD for your help. I have now tried the direct proof with the identity that martinbn suggested.I have now proceeded as follows:

$$=\sum\limits_{k=0}^{n}\frac{(-1)^k}{k+1} \binom{n}{k} x^{k+1}$$Then I formed the derivative with respect to x

$$=\sum\limits_{k=0}^{n} (-1)^k \binom{n}{k}x^k$$
$$=\sum\limits_{k=0}^{n} (-x)^k \binom{n}{k}$$
$$=\sum\limits_{k=0}^{n} (-x)^k 1^{n-k} \binom{n}{k}$$

Then I can use the binomial theorem, so the sum is ##(1-x)^n##

I then get the following equation ##(1-x)^n=\sum\limits_{k=0}^{n} (-x)^k 1^{n-k} \binom{n}{k}## which I integrate from 0 to 1:

$$\int\limits_{0}^{1} (1-x)^n \ dx=\int\limits_{0}^{1} \sum\limits_{k=0}^{n} (-x)^k 1^{n-k} \binom{n}{k} \ dx$$
$$\frac{1}{n+1}=\sum\limits_{k=0}^{n} \frac{(-1)^k}{k+1} 1^{n-k} \binom{n}{k}$$
$$\frac{1}{n+1}=\sum\limits_{k=0}^{n} \frac{(-1)^k}{k+1} \binom{n}{k}$$

Would that work as a proof?
 
  • #13
Lambda96 said:
Thank you PeroK, fresh_42 and WWGD for your help. I have now tried the direct proof with the identity that martinbn suggested.I have now proceeded as follows:

$$=\sum\limits_{k=0}^{n}\frac{(-1)^k}{k+1} \binom{n}{k} x^{k+1}$$Then I formed the derivative with respect to x

$$=\sum\limits_{k=0}^{n} (-1)^k \binom{n}{k}x^k$$
$$=\sum\limits_{k=0}^{n} (-x)^k \binom{n}{k}$$
$$=\sum\limits_{k=0}^{n} (-x)^k 1^{n-k} \binom{n}{k}$$

Then I can use the binomial theorem, so the sum is ##(1-x)^n##

I then get the following equation ##(1-x)^n=\sum\limits_{k=0}^{n} (-x)^k 1^{n-k} \binom{n}{k}## which I integrate from 0 to 1:

$$\int\limits_{0}^{1} (1-x)^n \ dx=\int\limits_{0}^{1} \sum\limits_{k=0}^{n} (-x)^k 1^{n-k} \binom{n}{k} \ dx$$
$$\frac{1}{n+1}=\sum\limits_{k=0}^{n} \frac{(-1)^k}{k+1} 1^{n-k} \binom{n}{k}$$
$$\frac{1}{n+1}=\sum\limits_{k=0}^{n} \frac{(-1)^k}{k+1} \binom{n}{k}$$

Would that work as a proof?
It's not wrong, but I'd say it's a bit ragged. You labour some points, like keeping the term ##1^{n-k}##, but skip any details of the integration entirely. The idea is simply$$(1 - x)^n = \sum\limits_{k=0}^{n} 1^{n-k}(-x)^k \binom{n}{k} = \sum\limits_{k=0}^{n} (-1)^k x^k \binom{n}{k} $$Hence:$$\int_0^1 (1 - x)^n dx = \int_0^1 \sum\limits_{k=0}^{n} (-1)^k x^k \binom{n}{k} dx = \sum\limits_{k=0}^{n} (-1)^k \binom{n}{k} \int_0^1 x^k dx$$Then do the integration.
 
  • Like
Likes Lambda96
  • #14
Thanks PeroK for looking over my calculation and thanks for the tip 👍👍, I have now written out my proof in more detail.
 
  • Like
Likes WWGD
  • #15
Lambda96 said:
Thanks PeroK for looking over my calculation and thanks for the tip 👍👍, I have now written out my proof in more detail.
By the way, the induction you tried unsuccessfully in your OP works quite well to prove the general result! Once you have proved the case for ##m = 1##, you can do an induction on ##m##. See post #11.
 
  • Like
Likes Lambda96
  • #16
I will try the proof via induction again, especially for the exam preparation :book:
 

FAQ: Induction with binomial coefficient

What is the binomial coefficient?

The binomial coefficient, denoted as C(n, k) or "n choose k," represents the number of ways to choose k elements from a set of n elements without regard to the order of selection. It is calculated using the formula C(n, k) = n! / (k!(n-k)!).

How do you prove properties of binomial coefficients using induction?

To prove properties of binomial coefficients using induction, you typically start with a base case to verify the property for a specific value, usually k = 0 or n = 0. Then, you assume the property holds for a general case k = m (inductive hypothesis) and prove it for k = m + 1. This step often involves algebraic manipulation and the recursive definition of binomial coefficients.

What is Pascal's identity and how is it related to induction?

Pascal's identity states that C(n, k) = C(n-1, k-1) + C(n-1, k). This identity can be proved using induction on n. By verifying the base case and using the inductive hypothesis, you can show that the identity holds for all positive integers n and k.

Can you use induction to prove the binomial theorem?

Yes, induction can be used to prove the binomial theorem, which states that (x + y)^n = Σ C(n, k) x^(n-k) y^k for k = 0 to n. The proof involves using the binomial coefficient properties and expanding the expression for (x + y)^(n+1) based on the assumption that the theorem holds for (x + y)^n.

What are some common mistakes when using induction with binomial coefficients?

Common mistakes include failing to properly establish the base case, incorrectly assuming the inductive step, and not correctly applying the properties of binomial coefficients. It's also important to ensure that all steps in the inductive proof are logically sound and follow from the inductive hypothesis.

Similar threads

Back
Top