Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Deduction of Gauss' Normal Distribution Function

  1. Oct 22, 2012 #1
    Hello, I'm David. I'm a new member here.

    Could anyone of you help me? Where can i find the formal deduction of Gauss' Normal Distribution Function? I've read a lot of statistics books and never found that. Where that comes from?

    It's just curiosity, not homework.

    P.S.: sorry about my english, I'm brazilian and my english is a little poor.
  2. jcsd
  3. Oct 22, 2012 #2

    Stephen Tashi

    User Avatar
    Science Advisor

    Instead of "deduction", you may mean "derrivation". A derrivation would be a theorem of the form "If .... then the distribution has the following formula...." and the formula would be the one that is used for the gaussian distribution. The most famous theorem of this type is the Central Limit Theorem http://en.wikipedia.org/wiki/Central_limit_theorem .

    If you are asking why the formula for the density for a gaussian distribution has the name "gaussian" or "normal", that's a historical question. There isn't any mathematical derrivation for it.
  4. Oct 23, 2012 #3
    Hi Stephen, thanks for the ansωer. I didn't knoω this theorem. Very interesting.
  5. May 7, 2014 #4
    Hi David.. I know this topic is very OLD (about 1.5 year), but when searching on google for exactly what you want, i was delivered to this forum... with no answer... so I kept searching and I found the derivation... you can find it here:


    P.S.: sorry about my english, I'm brazilian too.
  6. May 7, 2014 #5
    thanks, guicortei.
  7. Jul 5, 2015 #6
    The central limit theorem says that the arithmetic mean of a sufficiently large number of iterates of independent random variables will be approximately normally distributed,
    but that doesn't derive the normal distribution equation in general.

    And the pdf, just derives (again) an equation for a particular problem.


    But I can't demostrate is that in a experiment the measures follow that distribution.
  8. Jul 5, 2015 #7
    You can also derive the normal distribution as the limit of the binomial one.
  9. Jul 5, 2015 #8


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook