Proof
Let the sums Sp(n) be defined as
It follows from this that Sp(0+1)=Sp(0)+(0+1)p and so Sp(0)=0.
If the closed forms for Sp(n) can be expressed as polynomials they will not have constant terms, and will be of the form
for some matrix of coefficients ap,k. Substituting this into the inductive part of the definition above, applying the binomial theorem, and rearranging, we have
The double sum on the left hand side can be rearranged, noting that j≤k
Aligning coefficients then gives the relation
If p is a natural number (the reasoning so far has not assumed it is), then the right hand side is 0 when j>p, and we can assert that ap,k=0 for all k>p. Note that this is not the only possible solution; an infinite series that satisfied this property for its coefficients would also be valid, and would correspond to a closed form with a periodic component that was 0 at every natural number, such as sin(πn). The multiple solutions are an immediate result of the closed forms being defined only at the integers.
Asserting now that the matrix of coefficients is triangular, and multiplying both sides by j!, we have the recurrence relation giving all coefficients for any p (the first at j=p, the next at j=p-1, etc.).
which makes use of the Pochhammer symbol for the falling factorial.
Just as there is a diagonal relationship between binomial coefficients and between falling factorials, there is such a relationship between these coefficients, which can be demonstrated by 'breaking' the falling factorials at an arbitrary number t≤j.
Substituting k=k'+t and rearranging gives
This is of the same form as the original recurrence relation, which shows that
In particular, for k'=0, we have
The linear coefficients ap,0 are thus fundamental. These are the Bernoulli numbers, and so we have the final form
with each Bernoulli number derivable from this formula itself at n=1
and so B1=1/2 here.
Read more about this topic: Faulhaber's Formula
Famous quotes containing the word proof:
“The source of Pyrrhonism comes from failing to distinguish between a demonstration, a proof and a probability. A demonstration supposes that the contradictory idea is impossible; a proof of fact is where all the reasons lead to belief, without there being any pretext for doubt; a probability is where the reasons for belief are stronger than those for doubting.”
—Andrew Michael Ramsay (16861743)
“The thing with Catholicism, the same as all religions, is that it teaches what should be, which seems rather incorrect. This is what should be. Now, if youre taught to live up to a what should be that never existedonly an occult superstition, no proof of this should beMthen you can sit on a jury and indict easily, you can cast the first stone, you can burn Adolf Eichmann, like that!”
—Lenny Bruce (19251966)
“If any proof were needed of the progress of the cause for which I have worked, it is here tonight. The presence on the stage of these college women, and in the audience of all those college girls who will some day be the nations greatest strength, will tell their own story to the world.”
—Susan B. Anthony (18201906)