I would like to know if I can apply the Cauchy-Schwarz inequality to prove the following claim:
$$ \left( \int_0^\infty P(t) e^{-\theta t} t \ dt \right)^2 \leq \left( \int_0^\infty P(t) e^{-\theta t} \ dt \right) \left( \int_0^\infty P(t) e^{-\theta t} t^2 \ dt \right) $$ where $P(t)$ is a non-negative real function, $\theta$ is a positive scalar, and all integrals involved in the inequality converge.
My idea is to apply the Cauchy-Schwarz inequality (for non-negative integrands)
$$ \left( \int f g \right)^2 \leq \int f^2 \int g^2 $$
Setting $f= \sqrt{P(t) e^{-\theta t}}$ and $g = \sqrt{P(t) e^{-\theta t}} t$