Null hypothesis: OP will learn nothing from this thread

>>16913838
in a permutation order matters, in a combination it doesnt


$${}_{n}P_{r} = \frac{n!}{(n-r)!}$$

R is programming language for statistics

if someone offers you to pick one of three doors, kick him in the nuts and steal his wallet

Frogniggers are statistically low effort posters.

>>16913838
Statistically, there is a very high chance that a schizo has posted in this thread by the second post.

>>16913838
The variance of the asymptotic distribution of a sum of independent variables need not be the same as the asymptotic variance of the sum.

>>16913838
You can do arithmetic on random variables. Sometimes the resulting distributions even have closed-form formulas. For example, if you add two uniform distributions over the same interval you just get a triangular distribution.

>>16914284
>if you add two uniform distributions
two uniform random variables*

>>16914284
>if you add two uniform distributions
if you convolute two uniform distributions*

File: ICL.png (305.7 KB)

305.7 KB PNG

>>16913838
Poorly constructed models can cause great harm.

>>16913838
10 out of 10 pepefaggots are retarded nigger faggots, hope that helps

>>16914281
LMAO @ this NPD psycho still seething all this time later while every single thing I ever posted about here has been proven to be correct, while everything thing that comes out of his face continues, with 100% accuracy to be the stupidest shit you will hear that day.

You are a clown, a jester, sadly not even an amusing one, just one of the retarded ones people dont really laugh at, they just point to as a cautionary example of what total failure looks like. The fact that my mere existence makes you seethe uncontrollably brings me great joy

>>16914284
>>16914286
>>16914287
which one is it???

>>16914479
The first two are my post and my own correction. The third one is someone else but also true: the convolution of the PDFs of two independent random variables gives the PDF of their sum.

>>16914490
oic thank

>>16913838
Statistics fags have not discovered or created of note and have spent the last 2-3 decades leaching off machine learning chads and claiming they already knew whatever mlchads do.

>>16913838
half coins are neat!
https://en.wikipedia.org/wiki/Negative_probability
https://web.archive.org/web/20131108010759/http://www.wilmott.com/pdfs/100609_gjs.pdf

>>16913838
You can throw out probability and replace all the foundations of stats with game theory instead https://www.probabilityandfinance.com/

wtf why did you turn purple

>>16913838
The CRLB is actually only a special case of a much more general inequality known as the Van Trees/Lehmann Covariance Inequality. This inequality is a natural consequence of Cauchy-Schwarz and can be used to find statistical covariance bounds for arbitrary true sample and modeling distributions.

>>16914479
If [math] x_1 \sim p_1(x)[/math] and [math] x_2 \sim p_2(x) [/math] then [math] y = x_1 + x_2 \sim p_y = \int p_1(x)p_2(y-x)dx [/math]

>>16914284
nice

>>16914778
>muh ai
kys

>>16913838
Suppose X is a random variable, f(x) is the probability density function, and g(X) is an arbitrary function of X, then

[math] E(g(X)) = \int_R g(x) f(x) dx[/math]

>>16913838
If you fit a regression model for the purpose of inference, don't forget to validate model assumptions. Check QQ-plots, scale-location plots, etc. If you fit a GLM, check that the deviance does not vastly exceed the residual degrees of freedom. If you fit a GL(M)M, the DHARMa package in R is a surprisingly useful diagnostic tool.

>>16913838
"first moment" is just a fancy name for "expected value", which in turn is just a fancy name for mean value

>>16913838
23% of statistics are made up on the spot.

>>16919825
recent studies put it up to 37-42%