Riemann zeta function

[epkid08]

$$\zeta (s)= \frac{1}{(1-2^{1-s})} \sum_{n=0}^{\infty} \frac {1}{(2^{n+1})} \sum_{k=0}^{n}(-1)^k{n \choose k}(k+1)^{-s}$$

Is the main problem with trying to prove the hypothesis algebraically boil down to the fact that s is an exponent to a "k" term? Would a derivation of the function that had no k terms to an exponent s, be at all helpful to an algebraic approach to the hypothesis?

 

[epkid08]

This problem gives me a headache!