Evaluate $\sum_{k=1}^{n-3}\frac{(k+3)!}{(k-1)!}$.
My strategy is defining a generating function, $$g(x) = \frac{1}{1-x} = 1 + x + x^2...$$ then shifting it so that we get, $$f(x)=x^4g(x) = \frac{x^4}{1-x}= x^4+x^5+...$$ and then taking the 4th derivative of f(x). Calculating the fourth derivative is going to be a little tedious but it won't be as bad compared to the partial fraction decomposition I will end up doing. What is a better way to evaluate the sum using generating functions?
When you differentiate $g(x)$ $4$ times all powers of $x$ initially less than $4$ disappear anyway.
Then:
$$g^{(4)}(x)=4!(1-x)^{-5}=\sum_{k\ge 0}\frac{(k+4)!}{k!}x^k$$
Operating with $(1-x)^{-1}=1+x+x^2+x^3+\cdots$ on both sides gives a new expansion with terms which are partial sums of the coefficients of $x^k$ in the previous expansion:
$$(1-x)^{-1}g^{(4)}(x)=\sum_{r\ge 0}\left(\sum_{k=0}^{r}\frac{(k+4)!}{k!}\right)x^r$$
so
$$4!(1-x)^{-6}=\sum_{r\ge 0}\left(\sum_{k=0}^{r}\frac{(k+4)!}{k!}\right)x^r$$
$$\implies [x^r]4!(1-x)^{-6}=4!\binom{r+5}{5}=\sum_{k=0}^{r}\frac{(k+4)!}{k!}$$
but $r=n-4$ to match up with $n$ in the question, so
This summation is the same as the one in the question with only a shift in summation index.
Of course you may notice that this is just Pascal's hockey stick rule.