Skip to main content
Bumped by Community user
Bumped by Community user
Bumped by Community user
changed the title. It does not make much sense after the alternative formulations
Link

inequality involving exponential (looks simple)increasing functions

gave an alternative formulation that should be more interesting perhaps
Source Link

Let $a_k$ and $b_k$ be ascending positive numbers for $1\leq k \leq K+1$. If it is known that $$\frac{K\left(\exp\left(\frac{1}{K}\sum_{k=1}^K b_k\right)-1\right)}{\left(\sum_{k=1}^K \sqrt{a_k} \sqrt{\exp(b_k)-1} \right)^2}=1$$ I need to prove that $$\frac{(K+1)\left(\exp\left(\frac{1}{K+1}\sum_{k=1}^{K+1} b_k\right)-1\right)}{\left(\sum_{k=1}^{K+1} \sqrt{a_k} \sqrt{\exp(b_k)-1} \right)^2}<1.$$

From Matlab this surely holds, but I do not know why.

Here$\mathbf{\rm Alternative \;\,formulation.\;}$ Here is an alternative formulation of the same thing: With ascending $a_k$ and $b_k$, define $$z_m=\frac{\prod_{k=1}^{m}(1+b_k)^{(1/m)}-1}{\frac{1}{m}\left[\sum_{k=1}^m\sqrt{a_kb_k}\right]^2}.$$

Let $m_0$ be the smallest $m$ such that $z_m<1$. Then, for $m>m_0$, the sequence $z_m$ is monotonically decaying. It is easy to find counter examples where the sequence increase for $z_m$ with $m<m_0$.

$\mathbf{\rm 2^{nd}Alternative \;\,formulation, \;\,continuous\;\,version.\;}$ For positive monotonically increasing functions $A(x)$ and $B(x)$, define the function $$z(T) =\log\left(1+\frac{1}{T}\left[\int_0^{T} \sqrt{A(x)B(x)}\mathrm{d}x\right]^2\right)-\frac{1}{T}\int_0^T\log(1+B(x))\mathrm{d}x.$$

Let $T_0$ be the smallest $T$ such that $z(T)>0$. Then, $z(T)>0$ for $T>T_0$.

thanks

Let $a_k$ and $b_k$ be ascending positive numbers for $1\leq k \leq K+1$. If it is known that $$\frac{K\left(\exp\left(\frac{1}{K}\sum_{k=1}^K b_k\right)-1\right)}{\left(\sum_{k=1}^K \sqrt{a_k} \sqrt{\exp(b_k)-1} \right)^2}=1$$ I need to prove that $$\frac{(K+1)\left(\exp\left(\frac{1}{K+1}\sum_{k=1}^{K+1} b_k\right)-1\right)}{\left(\sum_{k=1}^{K+1} \sqrt{a_k} \sqrt{\exp(b_k)-1} \right)^2}<1.$$

From Matlab this surely holds, but I do not know why.

Here is an alternative formulation of the same thing: With ascending $a_k$ and $b_k$, define $$z_m=\frac{\prod_{k=1}^{m}(1+b_k)^{(1/m)}-1}{\frac{1}{m}\left[\sum_{k=1}^m\sqrt{a_kb_k}\right]^2}.$$

Let $m_0$ be the smallest $m$ such that $z_m<1$. Then, for $m>m_0$, the sequence $z_m$ is monotonically decaying. It is easy to find counter examples where the sequence increase for $z_m$ with $m<m_0$.

thanks

Let $a_k$ and $b_k$ be ascending positive numbers for $1\leq k \leq K+1$. If it is known that $$\frac{K\left(\exp\left(\frac{1}{K}\sum_{k=1}^K b_k\right)-1\right)}{\left(\sum_{k=1}^K \sqrt{a_k} \sqrt{\exp(b_k)-1} \right)^2}=1$$ I need to prove that $$\frac{(K+1)\left(\exp\left(\frac{1}{K+1}\sum_{k=1}^{K+1} b_k\right)-1\right)}{\left(\sum_{k=1}^{K+1} \sqrt{a_k} \sqrt{\exp(b_k)-1} \right)^2}<1.$$

From Matlab this surely holds, but I do not know why.

$\mathbf{\rm Alternative \;\,formulation.\;}$ Here is an alternative formulation of the same thing: With ascending $a_k$ and $b_k$, define $$z_m=\frac{\prod_{k=1}^{m}(1+b_k)^{(1/m)}-1}{\frac{1}{m}\left[\sum_{k=1}^m\sqrt{a_kb_k}\right]^2}.$$

Let $m_0$ be the smallest $m$ such that $z_m<1$. Then, for $m>m_0$, the sequence $z_m$ is monotonically decaying. It is easy to find counter examples where the sequence increase for $z_m$ with $m<m_0$.

$\mathbf{\rm 2^{nd}Alternative \;\,formulation, \;\,continuous\;\,version.\;}$ For positive monotonically increasing functions $A(x)$ and $B(x)$, define the function $$z(T) =\log\left(1+\frac{1}{T}\left[\int_0^{T} \sqrt{A(x)B(x)}\mathrm{d}x\right]^2\right)-\frac{1}{T}\int_0^T\log(1+B(x))\mathrm{d}x.$$

Let $T_0$ be the smallest $T$ such that $z(T)>0$. Then, $z(T)>0$ for $T>T_0$.

thanks

re-formulated the problem in an alternative fashion
Source Link

Let $a_k$ and $b_k$ be ascending positive numbers for $1\leq k \leq K+1$. If it is known that $$\frac{K\left(\exp\left(\frac{1}{K}\sum_{k=1}^K b_k\right)-1\right)}{\left(\sum_{k=1}^K \sqrt{a_k} \sqrt{\exp(b_k)-1} \right)^2}=1$$ I need to prove that $$\frac{(K+1)\left(\exp\left(\frac{1}{K+1}\sum_{k=1}^{K+1} b_k\right)-1\right)}{\left(\sum_{k=1}^{K+1} \sqrt{a_k} \sqrt{\exp(b_k)-1} \right)^2}<1.$$

From Matlab this surely holds, but I do not know why.

Here is an alternative formulation of the same thing: With ascending $a_k$ and $b_k$, define $$z_m=\frac{\prod_{k=1}^{m}(1+b_k)^{(1/m)}-1}{\frac{1}{m}\left[\sum_{k=1}^m\sqrt{a_kb_k}\right]^2}.$$

Let $m_0$ be the smallest $m$ such that $z_m<1$. Then, for $m>m_0$, the sequence $z_m$ is monotonically decaying. thanks It is easy to find counter examples where the sequence increase for $z_m$ with $m<m_0$.

thanks

Let $a_k$ and $b_k$ be ascending positive numbers for $1\leq k \leq K+1$. If it is known that $$\frac{K\left(\exp\left(\frac{1}{K}\sum_{k=1}^K b_k\right)-1\right)}{\left(\sum_{k=1}^K \sqrt{a_k} \sqrt{\exp(b_k)-1} \right)^2}=1$$ I need to prove that $$\frac{(K+1)\left(\exp\left(\frac{1}{K+1}\sum_{k=1}^{K+1} b_k\right)-1\right)}{\left(\sum_{k=1}^{K+1} \sqrt{a_k} \sqrt{\exp(b_k)-1} \right)^2}<1.$$

From Matlab this surely holds, but I do not know why... thanks

Let $a_k$ and $b_k$ be ascending positive numbers for $1\leq k \leq K+1$. If it is known that $$\frac{K\left(\exp\left(\frac{1}{K}\sum_{k=1}^K b_k\right)-1\right)}{\left(\sum_{k=1}^K \sqrt{a_k} \sqrt{\exp(b_k)-1} \right)^2}=1$$ I need to prove that $$\frac{(K+1)\left(\exp\left(\frac{1}{K+1}\sum_{k=1}^{K+1} b_k\right)-1\right)}{\left(\sum_{k=1}^{K+1} \sqrt{a_k} \sqrt{\exp(b_k)-1} \right)^2}<1.$$

From Matlab this surely holds, but I do not know why.

Here is an alternative formulation of the same thing: With ascending $a_k$ and $b_k$, define $$z_m=\frac{\prod_{k=1}^{m}(1+b_k)^{(1/m)}-1}{\frac{1}{m}\left[\sum_{k=1}^m\sqrt{a_kb_k}\right]^2}.$$

Let $m_0$ be the smallest $m$ such that $z_m<1$. Then, for $m>m_0$, the sequence $z_m$ is monotonically decaying. It is easy to find counter examples where the sequence increase for $z_m$ with $m<m_0$.

thanks

Source Link
Loading