hypothesis testing for exponential

jasper · **Joined:** Thu, 3 May 2012 20:45:29 UTC **Posts:** 2

Hi everyone! I'm having a lot of issues with this question. I have a bunch of data points and I want to prove that it came from an exponential distribution. Is the expected value for each category just 1/"sigma"? I did not receive a value for "sigma" - is that 1/the mean?

I am trying hard to find good examples to refer to, but am running into roadblocks.

Thank you so much.

mathematic · **Joined:** Fri, 1 Jul 2011 01:17:26 UTC **Posts:** 321

Usual terminology "sigma" is root mean square (square root of centralized second moment {variance}).

Shadow · **Posted:** Thu, 3 May 2012 21:58:01 UTC

jasper wrote:

Hi everyone! I'm having a lot of issues with this question. I have a bunch of data points and I want to prove that it came from an exponential distribution. Is the expected value for each category just 1/"sigma"? I did not receive a value for "sigma" - is that 1/the mean?

I am trying hard to find good examples to refer to, but am running into roadblocks.

Thank you so much.

$\sigma$ is the standard deviation.

jasper · **Joined:** Thu, 3 May 2012 20:45:29 UTC **Posts:** 2

Gah! I'm so sorry, I feel stupid now. I didn't mean "sigma," I think I meant gamma.

So according to this --http://www.aiaccess.net/English/Glossaries/GlosMod/e_gm_exponential.htm -- mu is equal to one over gamma. So would that mean the expected value of each frequency would be the same, one over mu?

outermeasure · **Posted:** Fri, 4 May 2012 11:46:15 UTC

Asssuming frequentist interpretation, just test goodness of fit as with any other distributions.

S.O.S. Mathematics CyberBoard

hypothesis testing for exponential

Who is online