0
$\begingroup$

I am unable to compute a complex integral which uses the pdf of log-normal distribution. Hence, I want to replace the pdf of log-normal distribution with an alternate function(s) (piece-wise approximation is also fine) $f_X$$(x)$ which is a function of $\mu$ and $\sigma^2$ so that it best approximates the pdf of log-normal distribution.

How can I approximate the pdf of log-normal distribution? Which function can I use?

P.S: $\mu$ and $\sigma^2$ are the mean and variance of the gaussian random variable using which the log-normal variable is derived.

$\endgroup$
0
$\begingroup$

You can try to approximate the pdf on a closed interval $[\mu - \alpha \sigma,\mu + \alpha \sigma]$ by an interpolation polynomial.

$\endgroup$
  • $\begingroup$ Can you please elaborate your answer by considering an example? $\endgroup$ – skt9 Jun 1 '16 at 10:10
  • $\begingroup$ @skt9 en.wikipedia.org/wiki/Lagrange_polynomial#Example_1 $\endgroup$ – Dark Jun 1 '16 at 14:04
  • $\begingroup$ Thanks, but I don't want to use numerical methods(interpolation etc), I would like to know if it is possible to replace the pdf by an alternative function under some conditions, like say a normal distribution when $\sigma$<<$\mu$(just an example, need not be true) or something like that? $\endgroup$ – skt9 Jun 2 '16 at 0:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.