The Student Room Group

Proof by Induction (Inequalities)

Hi,

I haven't come across proof by induction with inequalities before, so I'm not sure if my attempted method for this question is correct.

The question is:

(nN):n+12n(n+1)!(\forall n \in \mathbb{N}): n+1 \leq 2^{n} \leq (n+1)!

Here's how I proved it:

Basis Step:

n=1;(n+1=2)(2n=2)((n+1)!=2)n = 1; (n+1 = 2) \leq (2^{n} = 2) \leq ((n+1)! = 2)
 true for n=1\therefore\ \mathrm{true\ for}\ n = 1.

Assumption Step:

Assume true for n=kn = k
i.e. k+12k(k+1)!k+1 \leq 2^{k} \leq (k+1)!

Inductive Step:

With n=k+1n = k+1, the inequality becomes:
(k+1)+12k+1((k+1)+1)!k+22(2k)(k+2)!(k+1)+1 \leq 2^{k+1} \leq ((k+1) + 1)! \equiv k+2 \leq 2(2^{k}) \leq (k+2)!

Subtracting the original statement from the I.S.
(k+2)(k+1)2(2k)2k(k+2)!(k+1)!(k+2) - (k+1) \leq 2(2^{k}) - 2^{k} \leq (k+2)! - (k+1)!
12k(k+1)![k+21]1 \leq 2^{k} \leq (k+1)![k+2-1]
12k(k+1)!(k+1)1 \leq 2^{k} \leq (k+1)!(k+1)

We know kN\forall k \in \mathbb{N} that 1<2k1 < 2^{k}. We also know from the Assumption Step that 2k(k+1)!2^{k} \leq (k+1)!. As k1k \geq 1, (k+1)!(k+1)!(k+1)(k+1)! \leq (k+1)!(k+1) thus 2k(k+1)!(k+1)2^{k} \leq (k+1)!(k+1).

Conclusion Step:

Therefore, the inequality is true when n=k+1n = k+1. If the inequality is true for n=kn = k, then it is shown to be true for n=k+1n = k+1. As the result is true for n=1n = 1, it is now true nN\forall n \in \mathbb{N} by mathematical induction.
(edited 8 years ago)
Reply 1
Original post by r3l3ntl3ss
Hi,



Subtracting the original statement from the I.S.
(k+2)(k+1)2(2k)2k(k+2)!(k+1)!(k+2) - (k+1) \leq 2(2^{k}) - 2^{k} \leq (k+2)! - (k+1)!
12k(k+1)![k+21]1 \leq 2^{k} \leq (k+1)![k+2-1]
12k(k+1)!(k+1)1 \leq 2^{k} \leq (k+1)!(k+1)


I'm not convinced by this subtraction, though in this specific case it works, but in others the conclusion is not true, for example: 1351 \leq 3 \leq 5 and 2342 \leq 3 \leq 4
But 33≱213-3 \not\geq 2-1
Instead, I think you'll want to break it up into:
2k+1=2(2k)2(k+1)2^{k+1}=2(2^k) \geq 2(k+1) and a similar inequality on the right.
(edited 8 years ago)
agree with joostan

3 7 8

2 3 5

subtract

1 4 3 :redface:
With these kind of inequalities can you split them up? e.g first look at n+1 <= 2^n and then look at 2^n<=(n+1)! ?
Original post by mmms95
With these kind of inequalities can you split them up? e.g first look at n+1 <= 2^n and then look at 2^n<=(n+1)! ?


bump, does anyone know?
Original post by r3l3ntl3ss
Hi,

I haven't come across proof by induction with inequalities before, so I'm not sure if my attempted method for this question is correct.

The question is:

(nN):n+12n(n+1)!(\forall n \in \mathbb{N}): n+1 \leq 2^{n} \leq (n+1)!

Here's how I proved it:

Basis Step:

n=1;(n+1=2)(2n=2)((n+1)!=2)n = 1; (n+1 = 2) \leq (2^{n} = 2) \leq ((n+1)! = 2)
 true for n=1\therefore\ \mathrm{true\ for}\ n = 1.

Assumption Step:

Assume true for n=kn = k
i.e. k+12k(k+1)!k+1 \leq 2^{k} \leq (k+1)!

Inductive Step:

With n=k+1n = k+1, the inequality becomes:
(k+1)+12k+1((k+1)+1)!k+22(2k)(k+2)!(k+1)+1 \leq 2^{k+1} \leq ((k+1) + 1)! \equiv k+2 \leq 2(2^{k}) \leq (k+2)!

Subtracting the original statement from the I.S.
(k+2)(k+1)2(2k)2k(k+2)!(k+1)!(k+2) - (k+1) \leq 2(2^{k}) - 2^{k} \leq (k+2)! - (k+1)!
12k(k+1)![k+21]1 \leq 2^{k} \leq (k+1)![k+2-1]
12k(k+1)!(k+1)1 \leq 2^{k} \leq (k+1)!(k+1)

We know kN\forall k \in \mathbb{N} that 1<2k1 < 2^{k}. We also know from the Assumption Step that 2k(k+1)!2^{k} \leq (k+1)!. As k1k \geq 1, (k+1)!(k+1)!(k+1)(k+1)! \leq (k+1)!(k+1) thus 2k(k+1)!(k+1)2^{k} \leq (k+1)!(k+1).

Conclusion Step:

Therefore, the inequality is true when n=k+1n = k+1. If the inequality is true for n=kn = k, then it is shown to be true for n=k+1n = k+1. As the result is true for n=1n = 1, it is now true nN\forall n \in \mathbb{N} by mathematical induction.


I'm presuming this whole question is for n>0, just by the by.

I'd agree with your method up to just before you performed the subtraction thing, I think you've overcomplicated it after that. I'd solve this by just considering the two inequalities separately.

Firstly you've got:
k+22(2k)k+2 \leq 2(2^{k})
If we know that
k+12kk+1 \leq 2^{k}
I don't think it requires any more to notice that you're doubling the number on the right, and adding one to the number on the left. A quick check shows that for the number 1, adding 1 and doubling has the same effect, and for any other whole positive number doubling it will increase it more than adding one, which shows straight away that if this is true for n=k, it is also true for n=k+1

Secondly you've got:
2(2k)(k+2)(k+1)!2(2^{k}) \leq (k+2)(k+1)!
Again, we assume that
2k(k+1)!2^{k} \leq (k+1)!
and it should be obvious again just by looking at the equation, that the smallest effect multiplying by k+2 can have is to multiply by 2 (if k=0), but for any other k the effect of multiplying by k+2 is greater than the effect of multiplying by 2 for any whole positive number. So we again have shown that if this is true for n=k, it is also true for n=k+1 so have completed the induction.

There may well be a more rigorous 'mathsy' way to do this, but this is what my first thought was.
(edited 8 years ago)
Reply 6
Yeah I realised that it didn't work not long after haha. Thanks guys!

Quick Reply

Latest

Trending

Trending