SolAug1806
from Prof. Vasile Crîtoaje
Solution by landen with help from int-e
int-e helped a lot by pointing out how obnoxious this function is and that all the methods we had been trying wouldn't work. landen reluctantly got out Maxima (http://maxima.sourceforge.net). Somebody find a cute proof!
Proof: Since we have cyclic symmetry we can assume without loss of generality. Also notice that:
If , then . We want a lower bound for , it is sufficient to consider since the lower bound of will be a lower bound for the other permutations of . Sometimes this helps with cyclic symmetry.
Next, we, aka Maxima, homogenize:
Replace 1 by in the denominator to make the denominator a quadratic. Then multiply the numerator by to make it a quadratic also. Then get a common denominator. After this manipulation we get an equivalent problem:
This is a polynomial with every term of degree 6 and it is homogeneous which is usually a simplification of a constrained inequality. Next, we can set:
Maxima easily does this substitution and the result is: