2

It is given in the book by John B. Reade, titled: Introduction to Mathematical Analysis; on pg. 7; to prove the inequality

$$2^n \ge n^2, \ \ \forall n\ge 4, n\in \mathbb{N}.$$

It states that has distilled the proof using induction, & transitivity to the proof for the inequality: $2n^2\ge (n+1)^2, \forall n \ge 4$.

Am confused & feel that the proof before it for showing (Bernoulli inequality) $(1+x)^n\ge 1+nx$, by using induction, & transitivity must be applicable. Although, not clear how. Am presenting my attempt below based on that proof, but am unable to get the desired form of the inequality using that.

The inequality is true for the base case: $n=(n_0=)4$, i.e. $2^4 \ge 4^2$ is true.

Assume it is true for a particular $n\gt 4$ (inductive step), need show that it implies it is also true for $n+1$.

$2^{n +1} = 2^n.2\ge (n+1)^2\implies 2^n.2\ge (n+1)^2 = 2^{(n+1)}\ge (n+1)^2$. This can't be pursued further.

  1. Also, need show that $2^n.2\ge 2n^2$; but that leads to another proof using induction, & transitivity, as shown below. (I assume that the author takes this inequality to be known, hence not elaborates it).

For $n \ge 4, n\in \mathbb{N}$, need show $2^n.2\ge 2n^2$.
Base case $n=4: 2^4.2 \ge 2.16$.
Inductive case assumed for a value $n$.
Need show that $2^{(n+1)}.2\ge 2(n+1)^2\implies 2^{(n+1)}\ge (n+1)^2$

So, going nowhere.

jiten
  • 4,524
  • Possible duplicate of https://math.stackexchange.com/questions/1521611/how-can-i-prove-that-for-any-n-ge-4-2n-ge-n2 – VIVID Jul 18 '20 at 05:34
  • @Yourong'DZR'Zang I tried hard but could not infer anything except $\sqrt{2}n \ge (n+1)\implies \sqrt{2} \ge \frac{n}{n+1}$. Also, this expression is what want to prove above as derived finally. – jiten Jul 18 '20 at 05:35
  • @VIVID Seems to differ as no use made of $(1+x)^n\ge 1+nx$. But, the answers are good. – jiten Jul 18 '20 at 05:44
  • @VIVID If can show that the selected answer uses the property $(1+x)^n \ge 1+nx$, then proves my post too. – jiten Jul 18 '20 at 05:56
  • 1
    you mean you need a proof using that $(1+x)^n \ge 1+nx$? – VIVID Jul 18 '20 at 07:33
  • @VIVID Yes, thanks in advance. – jiten Jul 18 '20 at 07:41

1 Answers1

2

Clearly, we don't really need $(1 + x)^n \ge 1+nx$ to prove that $2n^2 \ge (n+1)^2, \forall n \ge 4$.

Since it is just saying that $$n^2 \ge 2n+1, \forall n \ge 4$$

which is equivalent to $$(n-1)^2 \ge 2, \forall n \ge 4$$

which we know it is true since if $n \ge 4$, the $(n-1)^2 \ge 9 \ge 2$.


If we insist to use $(1+x)^n \ge 1+nx$, we have

$$\left( 1-\frac1n\right)^2 \ge 1-\frac2n$$

Note that for $n \ge 5$, we have $1-\frac2n \ge 1-\frac25=\frac35 \ge \frac12 $.

Hence for $n \ge 5$, we have $\left(1-\frac1n\right)^2 \ge \frac12.$

which is equivalent to $$ 2(n-1)^2 \ge n^2, \forall n \ge 5$$

which is equivalent to

$$2n^2 \ge \left( 1+n\right)^2, \forall n \ge 4.$$

Siong Thye Goh
  • 149,520
  • 20
  • 88
  • 149
  • Please see my comment to the selected answer in the stated duplicate post (as by @VIVID) at: https://math.stackexchange.com/a/1521639/424260. I hope am correct, if not point out the fallacy - here, there, or in last chatroom at: https://chat.stackexchange.com/rooms/110379/new-room-108248. – jiten Jul 19 '20 at 07:00
  • Please see my today's post at: https://math.stackexchange.com/q/3767531/424260. It is for reason of inverse not existing, wrt a binop; yet two-sided identity existing. I mean why two-sided identity is not a sufficient condition for inverse to exist wrt a binop on a set. – jiten Jul 24 '20 at 08:08
  • Please respond to my last comment. I have stated after discussions that $x^2=e$ should be the condition, but it is NOT clear HOW is it possible for two-sided identity to exist; yet the property of $x^2=e$ not being followed. – jiten Jul 24 '20 at 10:13
  • I do not know what is your doubt, anyway, I need to go out to buy dinner before a contest in $45$ minutes. – Siong Thye Goh Jul 24 '20 at 10:15
  • I meant that if identity exists wrt a binop, then all non-identity elements ($x$) MUST be following the property $x^2=e$. I cannot think why this property is not being followed, yet identity exists. – jiten Jul 24 '20 at 10:16
  • Please see my edited OP. – jiten Jul 24 '20 at 10:42
  • I request an answer to the earlier stated post. – jiten Jul 25 '20 at 02:37
  • what is the question asking for? you already found the answer in a table right? – Siong Thye Goh Jul 25 '20 at 02:39
  • Yes, the table allowed to find out (no other commentor figured this out) that $x^2=e$ is the sufficient condition, while having double-sided identity is known a-priori as the necessary condition. But, what this lacks is that this is applicable only for groups of order $2$; as read by me on further reading. I have mentioned this in Edit $3$ in OP. I request an answer that covers order $2$ & higher order sets with binops to have condition to have inverse apart from two-sided identity.(I want to add that order of set is referring to their size in terms of no. of elements.) – jiten Jul 25 '20 at 02:50
  • i do not know the answer. – Siong Thye Goh Jul 25 '20 at 02:51
  • Can you please provide an incomplete answer that states my last comment here, so as to attract attention of others rather than getting deleted automatically. A new question with direct emphasis of such needed condition on sets with size $\ge 3$ is even more difficult to construct, & much more prone to negative votes. – jiten Jul 25 '20 at 02:56