4
$\begingroup$

Prove that $\frac{1}{2}\cdot \frac{3}{4} \cdots \frac{2n-1}{2n} \leq \frac{1}{\sqrt{3n+1}}$

I know this can be easily proved by induction. But I am looking for another approach. How do I prove this without induction? Here this question exists - How does one prove that $\frac{1}{2}\cdot\frac{3}{4}\cdots \frac{2n-1}{2n}\leq \frac{1}{\sqrt{3n+1}}?$. But the only one solution there uses induction. But I am looking for solution other than induction.

  • 4
    Possible duplicate of [How does one prove that $\frac{1}{2}\cdot\frac{3}{4}\cdots \frac{2n-1}{2n}\leq \frac{1}{\sqrt{3n+1}}?$](http://math.stackexchange.com/questions/119773/how-does-one-prove-that-frac12-cdot-frac34-cdots-frac2n-12n-leq)2017-02-09
  • 3
    @MyGlasses Nope... It is not asking to find a prove by induction .... The only answer there uses induction2017-02-09
  • 0
    Any general statement involving all natural numbers $n$ will at some level rely for proof on "induction" as an axiomatic principle that defines natural numbers. Perhaps you should explain more fully your motive in asking for a proof "without induction".2017-02-11

1 Answers 1

11

If we consider $$a_n = \frac{(2n-1)!!}{(2n)!!} = \frac{1}{4^n}\binom{2n}{n}=\prod_{k=1}^{n}\left(1-\frac{1}{2k}\right)\tag{1}$$ we have: $$ a_n^2 = \frac{1}{4}\prod_{k=2}^{n}\left(1-\frac{1}{k}+\frac{1}{4k^2}\right) = \frac{1}{4n}\prod_{k=2}^{n}\left(1+\frac{1}{4k(k-1)}\right)\tag{2} $$ hence: $$ 4n a_n^2 \leq \exp\sum_{k=2}^{n}\frac{1}{4k(k-1)} \leq \exp\left(\frac{1}{4}\right) \tag{3}$$ and: $$ a_n \leq \sqrt{\frac{1}{4e^{-1/4}n}} \tag{4}$$ is a stronger inequality, since $4e^{-1/4}\approx 3+\frac{1}{9}$.
No induction, just squaring and creative telescoping.

  • 0
    Wow, very nice! +12017-02-09
  • 0
    Seems to hard to me :( :'(2017-02-09
  • 1
    @RezwanArefin V. Hard. This is my attempt, hope it at least has some truth in it. http://imgur.com/a/KioSR2017-02-09
  • 0
    @Cppg Your's are harder than this one :V :V2017-02-09
  • 2
    This is so elegant ! Thanks for providing answers of this quality. Cheers.2017-02-09
  • 0
    Very nice solution! (+1). Reminiscent of your other solution [here](http://math.stackexchange.com/a/2122928/168053) but nice to see it in product form.2017-02-09