The Student Room Group

Econometrics

Just to give people an idea of what I'm currently doing and why I don't like econometrics...

Multivariate normal distribution

A k x 1 random vector X has the above said distribution with a mean vector mu and covariance matrix sigma that has the joint probability distribution function

f of x = 1 over root 2pi to k det sigma exp -1/2 x - mu multiplied by inverse of sigma multiplied by x - mu

Yep. That made loads of sense to me too.

Scroll to see replies

Reply 1
LOL I did that in high school!
Bring back some good memories:biggrin:
SWE_Ranger
LOL I did that in high school!
Bring back some good memories:biggrin:


You did matrix algebra back in high school did you? :wink:
Reply 3
Econometrics is probably on of the most important parts of any economics degree. good luck with it Ben:biggrin: .Enjoy:p:
Reply 4
sounds like a bit of S1 to me, shouldn't a casio graphical calculator be able to do all that easily?
Sorry - sigma is a covariance matrix. Not 'sum of'.

I'm too lazy to type it into bold... x and mu are vectors, one of which is primed. I had strangely assumed everyone would know this...!!

If you can do that from S1 knowledge, I'll give you a tenner. Better yet, what is the integral of f of x and why?
Reply 6
President_Ben
Sorry - sigma is a covariance matrix. Not 'sum of'.

I'm too lazy to type it into bold... x and mu are vectors, one of which is primed. I had strangely assumed everyone would know this...!!

If you can do that from S1 knowledge, I'll give you a tenner. Better yet, what is the integral of f of x and why?

You should've been paying attention in lectures, or attending them :wink:.
Reply 7
President_Ben
You did matrix algebra back in high school did you? :wink:

Actually I did, but that's still dutch to me. You make we want to not take econometrics. Well, except that ours is easier than yours :p:
Reply 8
President_Ben
Just to give people an idea of what I'm currently doing and why I don't like econometrics...

Multivariate normal distribution

A k x 1 random vector X has the above said distribution with a mean vector mu and covariance matrix sigma that has the joint probability distribution function

f of x = 1 over root 2pi to k det sigma exp -1/2 x - mu multiplied by inverse of sigma multiplied by x - mu

Yep. That made loads of sense to me too.


Ben, is this a compulsary module?
Reply 9
GarageMc
sounds like a bit of S1 to me, shouldn't a casio graphical calculator be able to do all that easily?


S1? Are you sure? It sounds a bit too advanced for AS Level statistics!
Reply 10
ba_ba1
Ben, is this a compulsary module?

I think so. UCL, and the University of London generally, goes quite hardcore on the Econometrics side. Plus, from the sound of it, I don't think Ben would take any more Econometrics than he needed to. :wink:
ba_ba1
Ben, is this a compulsary module?


Yep. Lucky us!!

Actually I did, but that's still dutch to me.


You should still be able to work out the integral of f of x and show why intuitively...

from the sound of it, I don't think Ben would take any more Econometrics than he needed to.


Got that right!! I could in theory spend all my final third year doing econometrics... but I think I want to pass my degree...
Reply 12
oooo soo looking forward to spending 1/3 of my 2nd yr on econometrics
Reply 13
President_Ben
Just to give people an idea of what I'm currently doing and why I don't like econometrics...

Multivariate normal distribution

A k x 1 random vector X has the above said distribution with a mean vector mu and covariance matrix sigma that has the joint probability distribution function

f of x = 1 over root 2pi to k det sigma exp -1/2 x - mu multiplied by inverse of sigma multiplied by x - mu

Yep. That made loads of sense to me too.

write it out properly (ie dont write f of x, write f(x) etc etc)
then i will find the integral maybe :p:
A k x 1 random vector, X has a multivariate normal distribution with a mean vector, mu and covariance matrix Sigma that has the joint probability distribution function

f(x) = 1/root[(2pi^k)detsigma]exp[-0.5(x - mu)prime Sigma^-1(x - mu)]

It's actually a very simple integration to do if you have a decent understanding of econometrics (but an explanation of why would be nice) and then an application of the above... better yet, a proof (but it's a very long proof...)
Reply 15
i feel ill
Reply 16
President_Ben
Just to give people an idea of what I'm currently doing and why I don't like econometrics...

Multivariate normal distribution

A k x 1 random vector X has the above said distribution with a mean vector mu and covariance matrix sigma that has the joint probability distribution function

f of x = 1 over root 2pi to k det sigma exp -1/2 x - mu multiplied by inverse of sigma multiplied by x - mu

Yep. That made loads of sense to me too.


What year are you in?
Reply 17
He is 2nd year at UCL.
Reply 18
i thought he was a lecturer? :confused: :p:
Reply 19
:p: No, he may just seem that way.

Quick Reply