MHB Generalized Holder Inequality: Proving the Inequality for Arbitrary Exponents

Click For Summary
The discussion focuses on proving the generalized Hölder inequality for arbitrary exponents, specifically addressing the condition where the sum of the reciprocals of the exponents equals one. The inequality states that the sum of the absolute values of the product of sequences is bounded by the product of the sums of the absolute values raised to their respective powers. A common approach to proving this inequality is through induction on the number of sequences involved. There is confusion regarding the necessity of approximating the exponents \( p_i \) by the form \( 2^N/n_i \). Overall, the thread emphasizes the mathematical framework and proof techniques related to the generalized Hölder inequality.
Mathkk
Messages
1
Reaction score
0
Let $a_i \in \mathbb R^n$ with $a_i = (a_{i}^j)_{j = 1 ... n} = (a_{i}^1, ... ,a_{i}^n)$ for $i = 1, ... , k$ and let $p_1,...,p_k \in \mathbb R_{>1}$ with $\frac1{p_1}+ ... + \frac1{p_k} = 1$

Then show the following inequality by assuming that there are for every $i = 1, ... ,k$ one $N \in \mathbb N_{>1}$ and one $n_i \in \{1,...,2^N\}$ with $p_i = 2^N / n_i$:

$$\sum_{j=1}^n|\prod_{i=1}^ka_{i}^j| \leq \prod_{i=1}^k(\sum_{j=1}^n|a_{i}^j|^{p_i})^{1 \over {p_i}}$$
 
Physics news on Phys.org
Mathkk said:
Let $a_i \in \mathbb R^n$ with $a_i = (a_{i}^j)_{j = 1 ... n} = (a_{i}^1, ... ,a_{i}^n)$ for $i = 1, ... , k$ and let $p_1,...,p_k \in \mathbb R_{>1}$ with $\frac1{p_1}+ ... + \frac1{p_k} = 1$

Then show the following inequality by assuming that there are for every $i = 1, ... ,k$ one $N \in \mathbb N_{>1}$ and one $n_i \in \{1,...,2^N\}$ with $p_i = 2^N / n_i$:

$$\sum_{j=1}^n\Bigl|\prod_{i=1}^ka_{i}^j\Bigr| \leqslant \prod_{i=1}^k\Bigl(\sum_{j=1}^n\bigl|a_{i}^j\bigr|^{p_i}\Bigr)^{1 / {p_i}}$$
The usual way to prove this is as a special case ($r=1$) of the more general inequality $$\Bigl(\sum_{j=1}^n\Bigl|\prod_{i=1}^ka_{i}^j\Bigr|^r\Bigr)^{1/r} \leqslant \prod_{i=1}^k\Bigl(\sum_{j=1}^n\bigl|a_{i}^j \bigr|^{p_i}\Bigr)^{1 / {p_i}},$$ where $\frac1{p_1}+ \ldots + \frac1{p_k} = \frac1r.$ That inequality is most easily proved by induction on $k$ (see the proof here).

I do not see how it helps to approximate the numbers $p_i$ by $2^N/n_i$. :confused:
 
Last edited:
We all know the definition of n-dimensional topological manifold uses open sets and homeomorphisms onto the image as open set in ##\mathbb R^n##. It should be possible to reformulate the definition of n-dimensional topological manifold using closed sets on the manifold's topology and on ##\mathbb R^n## ? I'm positive for this. Perhaps the definition of smooth manifold would be problematic, though.

Similar threads

  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 8 ·
Replies
8
Views
2K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 15 ·
Replies
15
Views
2K
  • · Replies 8 ·
Replies
8
Views
3K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 16 ·
Replies
16
Views
4K
  • · Replies 8 ·
Replies
8
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 2 ·
Replies
2
Views
3K