# Statistical inference ¶

Go back

You need to verify that the model hypothesis seems to be true. That's the hardest part, you will deal with estimators, likelihood, tests, ...

## Normal distribution ¶

According to the Central limit theorem (Théorème de la limite centrale)

$X \sim \lim_{n \to +\infty} N(\mu, \frac{\sigma}{n})$

You will use this a lot, because some tests are only working if the distribution is following a normal distribution, so you'll usually test that your distribution is following a normal distribution.