Proof Related to the Binomial Theorem

icestone111
Messages
10
Reaction score
0

Homework Statement


Use the above to prove that given a rational number a > 1 and A any other rational number, there exists b ε N such that ab > A.

Homework Equations


The above refers to the proving, by use of both induction and binomial theorem, that (1+a)n ≥ 1+na.

Binomial Theorem: (i=0 to n)Ʃ(n choose i)ai

The Attempt at a Solution



So I tried using the binomial theorem to get the value aN.
I get that aN must be greater than (i=0 to n-1)Ʃ(n choose i)ai
So how do I choose an N so that this holds?
Could you just let N > (i=0 to n-1)Ʃ(n choose i)ai?
 
Physics news on Phys.org
Let's keep it simple and use this result:
(1+a)n ≥ 1+na

The a in your problem and the a in the result they tell you to use are not going to be the same number. What should the relationship between them be?
 
Office_Shredder said:
Let's keep it simple and use this result:
(1+a)n ≥ 1+na

The a in your problem and the a in the result they tell you to use are not going to be the same number. What should the relationship between them be?


Thanks for the reply!

So, would the a in my problem be equivalent to (1+a), and then you would choose b = 1?
Then you would get (1+a)b ≥ 1+ba,
And using the expansion formula, they are equivalent when b = 1, so as a result b ≥ 1?
 
No, there's a difference between A and a. The objective now is to show that given any a>0, A>0, there exists some b such that (1+a)b>A. Use the fact that (1+a)b[/sub]>1+ab at this point
 
Sorry for the slowness...
It looks awfully similar to a Cauchy sequence proof.

But I do realize (?) that I am trying to solve for b in terms of the given/fixed values a and A.

When I look at the "hint" that was given, I keep seeing that (i=2 to b)Ʃ(b choose i)ai ≥ 0

Does this train of thought lead me anywhere?
 
Last edited:
Oh!
Would it work if we just choose b > (A-1)/a?

If it is I feel bad for missing something so simple. Looking too hard perhaps.
 
There are two things I don't understand about this problem. First, when finding the nth root of a number, there should in theory be n solutions. However, the formula produces n+1 roots. Here is how. The first root is simply ##\left(r\right)^{\left(\frac{1}{n}\right)}##. Then you multiply this first root by n additional expressions given by the formula, as you go through k=0,1,...n-1. So you end up with n+1 roots, which cannot be correct. Let me illustrate what I mean. For this...

Similar threads

Back
Top