-2

Hi it's an inequality found by chance using an integral analogue of the Kantorovitch inequality .

The problem :

$$300+\left(\int_{0}^{10}2-x^{-x}dx\right)\cdot\left(1+e^{-1}\right)^{-2}\cdot4\left(e^{-1}\right)>100\pi$$

I already know that the integral of $f(x)=x^{-x}$ between $0$ towards $\infty$ is less than $2$ and you can found other inequalities here Charming approximation of $\pi$: $2\left(\frac{1}{2}\right)^{\phi/2}+2< \pi$, where $\phi$ is the golden ratio and here Showing $\sqrt{\frac{e}{2}}\cdot\frac{e}{\pi}\left(\frac{e}{2}-\frac{1}{e}\right)<1$ without a calculator

Edit we can use the approximation :

$$2-x^{-x}\simeq 2-\left(g\left(x^x\right)\right)^{-1}$$

Where :

$$g(x)=1+\frac{\left(\ln\left(ax\right)\right)^{c}-\left(\ln\left(a\right)\right)^{c}}{\left(\ln\left(2a\right)\right)^{c}-\left(\ln\left(a\right)\right)^{c}}$$

Where we need some constraint as $g'(2)-1=0$ and $a=10000$

For further informations see Prove or disprove: $1+\frac{\ln^c(ax)-\ln^ca}{\ln^c(2a)-\ln^ca}-x\leq 0$, where $1<x$, $a>2$, and $c$ is chosen so that $f'(2)=0$

My question :

How to show it by hand without a calculator ?

Thanks.

Barackouda
  • 3,879

1 Answers1

4

Remark
Numerical calculation makes it seems true. \begin{align} 300+\left(\int_{0}^{10}2-x^{-x}dx\right) \cdot\left(1+e^{-1}\right)^{-2}\cdot4\left(e^{-1}\right) &\approx 314.15963 \\ 100\pi&\approx 314.15926 \end{align} But I doubt you would find this without use of calculator.

If you integrate to $9.9990$ and not $10$, then it is false.

GEdgar
  • 117,296