> Hello,
>
> We are students in American University of Sharjah (Marwan Aladdin and Hadi Aladdin)(Computer Engineers).. and we are working on a new Math Formula.
>
> Our current research is in "Number Theory". We found a relation between Prime Numbers
What do you mean by "relation between Prime Numbers"?
> and we came up with a formula to GENERATE ANY PRIME NUMBER USING ITS N WITHOUT USING >DATABASES OR COMPARISON!!
There are many such formulae. They are well known. Read, for
example:
P. Ribenboim "The Book of Prime Number Records".
However, such formulae are computationally useless for large N.
> And this works for any prime even in millions. Currently we need super computers to do Regression Fitting and >Curve Fitting for our formula to make it work for HUGE n values..
Supercomputers are not needed to do curve fitting. And trying to
compute primes via a fitted
curve is useless once the prime to be computed is outside the range of
your fitted curves.
A sieve would be much more efficient.
>or in other words we need a database of functions and search the best fit function and its regression.
>
> What do you suggest to achieve this?
Don't. You are wasting your time.
>Is there any special way for such researches? Any known research centers who can help do this task? Any >special software suggested to be installed on super computers to achieve this?
I suggest that you do some background reading. Otherwise, you are
proceeding without really knowing the required background mathematics.
Start with the book I mentioned above. You should
also look at:
Crandall & Pomerance: "Prime Numbers; A computational Perspective"
R. Guy, "Unsolved Problems in Number Theory"
H. Riesel "Prime Numbers and Computer Methods for Factorization".
raladin - 24 Jul 2007 13:08 GMT
Thanks for your reply Pubkeybreaker.
What we are trying is not to plot the curve of consecutive prime numbers..according to some confidential information, we can't provide you with the method.. but have done a method on prime numbers, and we got a relation (a Non-Linear Perfect Exponant Relation). We tested it for the 26,000 nth prime, and the error % was "0.2%". What we are trying is to make the error much less using Series. I hope you get the idea more clearer now..
We have checked the Reimann hypothesis and other trials of other scientists for Prime Numbers Distribution.. but couldn't find a formula for generating nth Prime Number (without knowing the n-1 primes).
Do you get it clearer now maybe?
So what do you suggest? :-)
Thanks alot for your great help in advance! We do really appreciate your time and effort..