/sci/ - Science & Math

>>8737595
thats what you get for selling us out, you piece of shit. enjoy your pile of money, brainlet

>>8737595
>where's the fucking beauty
in the salary

>>8737600
SHUT UP SHUT UP

>>8737595
It's not all that arbitrary, you're just comparing real data to distributions for the most part, and seeing how they match.

>>8737595
You sold your soul for a paycheck.

Hope you've got big plans for all that money, because that's all you have left.

>>8737600
GAS THE BRAINLETS
IQ WAR NOW

>arbitrary
brainlet detected

>>8737595
Sssssssssell Ouuuuuuuuut
Sssssssseeeeeeeell
Ouuuuuuuuuut
You have sold out you're own kiiiind~
Hisssssssssssss~
Hissssssssssssssss~

>>8737595
enjoy your long career and high income fucking loser

>>8737595
Here's an assignment, kid: calculate the probability that you'll kys yourself before graduating.

>>8737595
>tfw planning to do math undergrad and masters in stats

>>8737595
Go further OP. I have a PhD in Biostatistics and I love what I do (and the fat 5 figure paycheck I get every month)! I also get to make actual contributions to society that won't be obsolete in 40 years.

That being said, if you're doing a lot of memorizing and find what you're studying to be inelegant, you may be doing it wrong... Statistics is just as elegant and rigorous as any other field in Mathematics. Mathfags just like to say otherwise because they jelly we make more than them and know just as much math except for that one tiny subject they spent the last 20 years sperging out over.

>>8737595
I want to run a regression analysis with three outputs.

Essentially I'm looking at roads as output, with three different categories of road.

All I know how to do is single output regressions please help Anons.

>>8739892
>Satan only paid him 5 figures for his immortal soul

>>8741944
>Thinks 5 figures per month translates to 5 figures per year.

Poor anon...

>>8741937
I'm gonna go ahead and assume the categories of road have no order associated with them. If that's the case, use logistic regression with a g-logit link function. This is commonly referred to as a baseline logistic model and it is not difficult to interpret.
Basically, you pick a category of road to be the reference category, and you can simultaneously fit two separate logistic regression equations to model the odds of being in each of the other two road categories as opposed to the reference road category.
In essence you are given two separate logistic regression equations for each of the two road categories being compared to the reference road category, and these models can be interpreted in the usual fashion. The difference in parameter estimates can also be used to compare the two non-reference categories.

>>8737595
Since you've completed a math undergrad, I assume you've taken real analysis. Look functional analysis for all theory behind almost everything done in mathematical statistics. If you've taken real analysis, you should be equipped for everything talked about in functional analysis. Don't lose sight of the forest for the trees. There's beauty behind the calculations. Most places don't bother teaching it though. You have to search.

>>8739892
you're doing what I want to do anon, how long did it take for you to do your PhD?

also what does your job look like?

>>8737595
>statistics
>where's the fucking beauty
You went into the wrong branch for that, dump frogposter.

>Beauty

Math is just a tool to be used by real scientists to do things that have actual consequence. You're just studying a different set of tools; that's all.

>>8743295
5 years. Fairly standard for stats PhD.

Generally, I help design and conduct clinical trials (primarily phase 2/3). I'm rather newly minted, so there's a more senior statistician that oversees most things, but we'll all get there one day. A decent portion of my week is spent in meetings providing statistical consulting to MDs and lab scientists. Much of my remaining time is devoted to grant writing and overseeing of general trial operations as well as to methodological research.
Clinical trials are a huge ordeal, and I'd say the primary focus of my job is to ensure the trial is designed to statistically address the question physicians want answered and to ensure it does so while placing a minimal burden on patients.
There are also plenty of other faculty in my department who's research is more related to fMRI/EEG/Omic analysis, so I'd say it's a pretty broad field in which you should be able to find something that interests you. It all often overlaps anyways.

BAYESIAN
A
Y
E
S
I
A
N

>>8743771
absolutely based

Welcome to the Alabaster Tower OP

>>8743771
I want to pursue a degree in economic statistics, this post was really encouraging, thank you anon

Masters programs are for brainlets who lack confidence. Straight to phd or kys