I was just wondering if there are any other Statistics guys here. And if not what are your opinions of Statistics in general.
My MS was Math but this semester I started doing statistics and it's really nice. I never wanted to be a math theorist, but I'd love doing some mathematical stats and have a strong understanding of Probability.
Not a stats major, but I'm a CS major with a math/stats minors. Currently taking a graduate pattern recognition course. Hol-ee fuck. I did not imagine a graduate pattern recognition course is basically a graduate probability course with a ton of analysis. It is really fun though, to learn the background of a bunch of ML algorithms from a statistics point of view rather than CS point of view for a change.
However, I love statistics and probability. They're so useful and a lot of the time they're intuitive. It's going to be an increasingly growing field in the near future but it's off most people's radars currently. Stats majors are basically for the really smart weird kids that already know they are going/have to go to graduate school because every stats department in the nation is really small so they end up taking all the undergrad course rather quickly.
Word.
I always wondered about the usefulness of a Stats BS (on its own).
Most people who do stats for grad school did their BS in Math. And if you look at the history of statistics it’s mostly guys with math degrees except for Casella. You need that real math background to do higher level stats.
But it’s a great route to take if your career goal isn’t “getting published in Journal if Complex Dynamical Systems while grading kids’ Cal2 homework for the rest of your life”
>It's going to be an increasingly growing field in the near future but it's off most people's radars currently.
Dude Data science is the biggest and fastest-growing part of STEM for the last decade.
data science != statistics and probability
Half of those "data scientists" don't know the probability of an event is the integral on a continuous probability. They probably don't know parameter estimation or any distribution besides Gaussian.
on a continuous distribution*
Stats is where it's at, but if you cannot code then you'll never make it.
I can't do math to save my life. But my dad is a statistician and has had a very successful career doing clinical trials for pharmaceutical companies and he loves it. He says he genuinely enjoyed what he does and likes his work.
>If you cannot code
I refuse to believe there exists someone who can't learn to code
Let me introduce you to the average girl/guy and their boomer parents
Any data scientist from a non-shit school will know these things. They're literally freshman level prob&stats
Another benefit of learning statistics: you realise that the majority of natural and social scientists have absolutely no idea what they are doing in experimental design and statistical analysis and you realise that most of those scientists are those same brainlets who chose to major in a non-mathematical science because they found math too hard, and that this is the fundamental reason why there is a crisis of replication in psychology, biology and medical science.
[Judges you in Bayesian]
I'd say a good 40% of guys can't code for shit. Double that for women. Add 20% if normie.
Who /statistigrad/ here?
Whats the second most useful distribution to familiarize myself with after I'm comfortable with the Gaussian?
...
god damn thats nifty
god damn whats that from ?
If you do data or network science, then be familiar with the power law/pareto distribution (as the distribution of degrees of vertices in real-life graphs usually follow a power law)
ok so I'm testing nootropics on myself and if I test for five days, doing baselines before each dose how would I check if my baseline increased? considering it can go up and down based on so many factors I was thinking take the last four days, add them all up, divide by 4 then compare that number to the baseline from day 1 (before any dose), is this a good way to do it?
>(as the distribution of degrees of vertices in real-life graphs usually follow a power law)
lmao of course they do
These kinds of diagrams aren't uncommon.