In lecture, we were asked:
Does $ \sum \limits_{n=0}^\infty \frac{ 2n^3+ 3n -8 }{ n^5-5n^3-n^2+2 }$ converge?
$a_n = \frac{ 2n^3+ 3n -8 }{ n^5-5n^3-n^2+2 }$
In discussing strategies for applying a comparison test for $a_n$ we rejected several options for choosing a bound for $b_n$. The strategy is to choose an expression larger in the numerator and smaller in the denominator for $b_n$ If I understand correctly, using $2n^5$ in the denominator for $b_n$ doesn't work, because, as n gets large, we have
$$ \frac{2n^3}{n^5} \not < \frac{3n^3}{2n^5}$$
$$ 2 \not < \frac{3}{2} $$
If this is the case, couldn't we choose $b_n = \frac{10n^3}{2n^5}$?
Plugging the first inequality into Wolfram Alpha was not enlightening.

From the lecture:
To apply the comparison test, we need to bound the numerator from above and the denominator from below.
For the numerator of $b_n$, $2n^3$ won't work because we have $3n$ in the numerator of $a_n$. $3n^3$ is large enough.
$2n^5$ in the denominator of $b_n$ won't work.
$a_n = \frac{ 2n^3+ 3n -8 }{ n^5-5n^3-n^2+2 } \ \substack{ <\\ \\ >} \ \frac{3n^3}{2n^5} =b_n$
Using $n^4$ in the denominator won't work because this gives us a harmonic series for $b_n$, which is divergent.
$a_n = \frac{ 2n^3+ 3n -8 }{ n^5-5n^3-n^2+2 } \ \substack{ <\\ \\ >} \ \frac{3n^3}{n^4}=b_n $
Using asymptotic analysis, we get $a(x)= \frac{2}{x^2} + \mathcal{O}(\frac{1}{x^4})$, which gives us an integral with p=2, which converges.