Optimizing Inequality with Constraints

  • Thread starter Thread starter nasshi
  • Start date Start date
  • Tags Tags
    Inequality
AI Thread Summary
The discussion revolves around proving an inequality under specific constraints, with the goal of personal gain rather than academic purposes. The initial inequality is challenged with the example (a,x,y,z)=(1/2,2,1,1), which is deemed incorrect. Adding the constraint y+z>x leads to a successful proof using the values (a,x,y,z)=(1/2,2,1,2). However, concerns are raised about the effectiveness of the method for smaller values of x, y, and z, prompting a request for additional constraints to solidify the proof. The conversation highlights the complexity of establishing a general proof while navigating the limitations of the chosen constraints.
nasshi
Messages
18
Reaction score
0
I'm interested in proving the following inequality, given the following constraints. This is for personal gain, and not for homework. I'm not even sure if I can prove this without additional constraints. If you can solve it by adding constraints, please let me know what options I have for adding constraints.

Given:
0<a<1
x>0, y>0, z>0
x>y

Prove:\frac{x-y}{(1+x)(1+y)} &lt; \frac{z*a}{1+y+z}

Thanks so much for your help!
 
Last edited:
Mathematics news on Phys.org
I believe it's proven wrong with the example, (a,x,y,z)=(1/2,2,1,1).
 
You are indeed correct.

How about we add the constraint:
y+z&gt;x

What about now?
 
Well y+z>x really just means z>x in your case, but I'll try to prove it:

<br /> \frac {x-y}{(1+x)(1+y)}&lt;\frac {z*a}{1+y+z}

Fallowing constraints, we set (a,x,y,z) to (1/2,2,1,2)

<br /> <br /> \frac {2-1}{(1+2)(1+1)}&lt;\frac {2*1/2}{1+1+2}

\frac {1}{6}&lt;\frac {1}{4}

This proves true.
Assume (a,x,y,z)\longrightarrow(a+1,x+1,y+1,z+1); plug in the original set.

\frac {(x+1)-(y+1)}{(1+(x+1))(1+(y+1))}&lt;\frac {(z+1)(a+1)}{1+(y+1)+(z+1)}

\frac {x-y}{(x+2)(y+2)}&lt;\frac {(z+1)(a+1)}{y+z+3}

Now plug in, and simplify:

\frac {2-1}{(2+2)(1+2)}&lt;\frac {(2+1)(1/2+1)}{1+2+3}

\frac {1}{12}&lt;\frac {9}{12}

I don't know if this is formal, but I'd call it adequate.
 
I think it works well for larger values of x,y,z. But when x,y,z are small, for example, between 0 and 1, I think your method won't work.
 
I should have set (a,x,y,z) to (d,2d,d,3d) and (1-d,2d,d,3d).
Anyways, you have to fix your constraints...
 
I can't exactly "fix my constraints". I'm trying to prove something as generally as possible. Can you think of any constraints I can add that would easily establish the desired result? Because I can't.
 
Here are some constraints that I think work:
x>y
0<y<1
a>6
z>2/35
 
Back
Top