Can Hölder's Inequality also be applied to other optimization problems?

  • Context: MHB 
  • Thread starter Thread starter anemone
  • Start date Start date
Click For Summary
SUMMARY

Hölder's Inequality can be effectively applied to optimization problems, specifically in proving that if \(c^2 + d^2 = (a^2 + b^2)^3\) for positive \(a, b, c, d\), then \(\frac{a^3}{c} + \frac{b^3}{d} \ge 1\). The discussion highlights the use of Lagrange Multipliers as a powerful tool for finding extrema in optimization scenarios. A proposed solution demonstrates the transformation of variables \(x\) and \(y\) to leverage Hölder's Inequality, leading to the conclusion that \(x^{\frac{3}{2}} + y^{\frac{3}{2}} \ge 1\).

PREREQUISITES
  • Understanding of Hölder's Inequality
  • Familiarity with Lagrange Multipliers
  • Basic knowledge of algebraic manipulation
  • Concept of optimization in mathematical analysis
NEXT STEPS
  • Study the applications of Hölder's Inequality in various optimization problems
  • Learn about Lagrange Multipliers and their role in constrained optimization
  • Explore advanced algebraic techniques for manipulating inequalities
  • Investigate other inequalities used in optimization, such as Cauchy-Schwarz Inequality
USEFUL FOR

Mathematicians, students studying optimization techniques, and anyone interested in applying inequalities to solve mathematical problems.

anemone
Gold Member
MHB
POTW Director
Messages
3,851
Reaction score
115
If $a, b, c, d >0$ and $c^2+d^2=(a^2+b^2)^3$, then show that $$\frac{a^3}{c}+\frac{b^3}{d} \ge 1$$.
 
Physics news on Phys.org
Re: Prove (a³/c)+(b³/d) ≥1

My solution
This probably can be done with some clever manipulation of inequalities but the problem does suggest that the method of Lagrange multiplier would be effective.

If we define

$F = \dfrac{a^3}{c} + \dfrac{b^3}{d} + \lambda\left(c^2+d^2-(a^2+b^2)^3 \right)$,

so
\begin{eqnarray}
F_a &=& 3\,\frac {a^2}{c}-6a\,\lambda \left( a^2+b^{2} \right) ^{2}\\
F_b &=&3\,\frac {b^2}{d}-6b\,\lambda \left( a^2+b^{2} \right) ^{2}\\
F_c &=&-\,\frac {a^3}{c^2}+2\,\lambda c\;\;\;\;\;(*)\\
F_d &=&-\,\frac {b^3}{d^2}+2\,\lambda d\\
F_{\lambda} &=&c^2+d^2-(a^2+b^2)^3
\end{eqnarray}

Setting $F_c=0$ and solving for $\lambda$ gives $\lambda= \dfrac {a^3}{2c^3}$. From the $F_a=0$ equation we have

$-3\,{\dfrac {{a}^{2} \left( -c+{a}^{3}+a{b}^{2} \right) \left( c+{a}^{
3}+a{b}^{2} \right) }{{c}^{3}}}
=0$

from which we obtain $c = a^3+ab^2$. Factoring the $F_b=0$ equation gives

$3\,{\dfrac { \left( {a}^{2}b+{b}^{3}-d \right) b}{ \left( {a}^{2}+{b}^{
2} \right) d}}=0$

leading to $d = a^2b+b^3$ noting that the remaining equation in (*) are satisfied. Thus, we have

$\dfrac{a^3}{c} + \dfrac{b^3}{d} = \dfrac{a^3}{a^3+ab^2} + \dfrac{b^3}{a^2b+b^3} = 1$

the minimum as required.
 
Re: Prove (a³/c)+(b³/d) ≥1

Thanks for participating, Jester and well done! And now it seems to me Lagrange Multiplier is a magic tool and can be applied to all kind of optimization problems to find the desired extrema points.

I want to share with you and others the solution proposed by other as well. It suggests the use of Hölder's Inequality to solve this problem.

If we let $$x=\left(\frac{a^3}{c} \right)^{\frac{2}{3}}$$ and $$y=\left(\frac{b^3}{d} \right)^{\frac{2}{3}}$$, we see that

$$a^2+b^2=c^{\frac{2}{3}}\left(\frac{a^3}{c} \right)^{\frac{2}{3}}+d^{\frac{2}{3}}\left(\frac{b^3}{d} \right)^{\frac{2}{3}}$$

$$\;\;\;\;\;\;\;\;\;\;\;\;=c^{\frac{2}{3}}(x)+d^{ \frac{2}{3}}(y)$$

$$\;\;\;\;\;\;\;\;\;\;\;\;\le (c^2+d^2)^{\frac{1}{3}}(x^{\frac{3}{2}}+y^{\frac{3}{2}})^{\frac{2}{3}}$$

$$\;\;\;\;\;\;\;\;\;\;\;\;\le (a^2+b^2)(x^{\frac{3}{2}}+y^{\frac{3}{2}})^{\frac{2}{3}}$$

Hence

$$a^2+b^2\le (a^2+b^2)(x^{\frac{3}{2}}+y^{\frac{3}{2}})^{\frac{2}{3}}$$

$$x^{\frac{3}{2}}+y^{\frac{3}{2}} \ge 1$$

$$\therefore \frac{a^3}{c}+\frac{b^3}{d} \ge 1$$ (Q.E.D.)
 

Similar threads

  • · Replies 2 ·
Replies
2
Views
4K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 17 ·
Replies
17
Views
2K
  • · Replies 7 ·
Replies
7
Views
4K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 7 ·
Replies
7
Views
2K
  • · Replies 21 ·
Replies
21
Views
3K
  • · Replies 0 ·
Replies
0
Views
2K