hey Veeky Forums lets talk about math jobs for those of us who have tread down the applied math path.
we'll take economics, statistics, physics, and comp sci in our conversation too.
what's popular these days? what are you currently doing? what do you want to do?
post your
age/degree/school/job
>22, applied mathematics, Berkeley
>interning at a bank in SF doing statistical stuff. (quant)
although trying to get into software engineering, i feel thats where the fun is.
Other urls found in this thread:
ams.org
twitter.com
also, pic semi-related. walter rudin.
>22
>Bachelor's in Mathematics
>University of [X]
>Software Developer
>what's popular these days?
Considering the fact that some days ago my fucking old as fuck dad sent me an email with a blogpost about Data Science, it looks like Data Science is the meme of the month. I suppose.
It is unrelated, but data science looks like fucking shit.
>what are you currently doing?
Software Development. It is comfier than I thought it would be because of the scope of the projects I work 100% alone and I only have to report to clients, but clients cannot talk directly to me so I live in a cloud where I just have to casually think about work.
Use mostly Java and VB.Net so I'm not exactly living the C++ dream but with Java I can pretend.
I basically construct simple algorithms that just have a lot of moving parts and don't have inherent complexity in them, but still these are fun problems to think about and like the pretend mathematician I am, I love it. Then I have to design UIs because text interfaces are not acceptable in 2016 but that is fine too because it is quick and not too boring.
All that I do is related in some way or another to finance.
>what do you want to do?
'Graduate' from finance to science/math but honestly I don't care that much. Maybe I will one day do a MS in something related and then live the dream.
>although trying to get into software engineering, i feel thats where the fun is.
It is okay. I enjoy a lot but I feel like if I told you it is amazing I would be lying to you. It is just 'ok'.
Oh, and I want to add this and I hope others will agree.
There is a metric on payscale that says the percentage of people from every career who believe that they are making the world a better place with their careers.
Please answer that too, if you think you are doing good.
I'd say yes. Because of me, entrepreneurs looking to get loans to fund their cancer research company can calculate the cost of their payment more comfortably and faster.
>math
>jobs
HAHAHAHHAHAA
if you want a job get an engineering degree
im actually in a data science organization at Berkeley, it's not too bad actually.
machine learning and deep-dive neural networks are doing some really cool stuff.
you should look into it more.
>kek, my dad also sent me blog posts about Data Science.
it's what my applied math concentration is in btw.
>do i think i'm doing the world good?
not at all. i'm helping the rich get richer.
im quite easily employed with my applied math degree. although it is from a top 5. so there is that.
you know there's some websites where you get paid to do people's math homework.
i wonder if Veeky Forums knows god ones. theres transtutor i've seen, idk if its good or not
>im actually in a data science organization at Berkeley, it's not too bad actually.
That is interesting.
My question to you is
>Do you or does your team program their own algorithms or do you just do scripting on top of pre-existing algorithms?
If you do your own algorithms:
>How deep does this go? At what level do you start using libraries instead of your own code.
If you just do scripting on top of an existing codebase:
>Is this the biggest chunk of your work day or is there another part of it that is the main thing you do?
Mathematics has lower unemployment than most fields, including some engineering. Nobody actually bothers to use facts or statistics.
>25, pure math, UT
>DB management (admin until I get experience/training)
I want to believe.
What does your work day entail and where do you use math and logic?
definitely will use libraries whenever we can. I use numpy and datascience (for python), no sense in fixing what isn't broken. that's for the analysis portion.
>do we use our own algorithms.
yeah, in part.
the last project that I was on was a search algorithm, and it needed modification for a new feature. we looked at the data, decided how to implement said feature, then wrote the code for what we decided would be best.
we were given his existing codebase, and an assload of logged data.
we pushed our code into the website, then tested it for a week to see improvement on the visitors, then rinse and repeat.
the biggest part? hmm, not sure, everyones doing different thing. My part was analyzing the data, and how to build the algorithm, i didnt work with implementation.
It's a civil engineering company. I started out doing data analysis. I now manage our databases since I've had training under our current DB admin.
Work day isn't all that much. Without getting too much into detail about the type of firm, we do create DB's for new projects and add them to our servers depending on the clients requested work.
Databases we use SQL obv, but we're implementing our new (proprietary) software to channel through our DB's smoothly and efficiently. With Microsoft SQL 2016 coming out, we're hoping to get our company to get it since it'll be soooo much easier and better for us.
I'm not coding all day. Since it's a small firm, half the time I'm their IT guy.
To answer your question more specifically I use my math/logic in the scripts I write. Although it isn't directly math related like group theory or proving l'hopitals rule, we do use some multivariate statistical functions as well as some engineering equation.
can someone explain to me why "machine learning" is so cool? i looked up some and it just looked like basic optimization, numerical stuff, and statistics phrased in a way that compsci majors would understand
its the future, man. basically it's the process of creating an algorithm and automating it to train itself on a set of sample data, so that you can use it in the real world.
e.g. Baidu had a problem they couldn't solve. They wanted to program a remote controlled hellicopter to do ariel acrobatics. However, their best programmers could not do this. too many vectors, too many signals, too much noise.
so they developed an AI which watched thousands of hours of flight data from professional pilots.
the AI taught itself how to fly the helicopter with ariel acrobatics.
ie. the AI wrote a program that our best programmers could not.
>28, pdhd of methmatics
>suicide after a decade of debt slavery
:|
theres some other cool stuff. i heard them at a conference hah.
29 / PhD Mathematics / Carnegie Mellon / Data Scientist
The job is a blast, I work at a startup doing research
Mfw the high school math teachers here all have engineering degrees. Kek.
This guy gets it
>i use python numpy
>24
>Bachelor's: Pure math: Pitt
>Master's in math: focus in statistics/ ML: University of Mississippi
(Was a PhD student, but wanted out for a while)
>RND analyst/ Data Analyst at a major insurance company
Just started my job. I'm enjoying it thus far; they are thrilled that I have such a broad mathematical background/ encourage me to think of creative ways to implement the stuff I've learned.
Insurance is far more complex than I used to naively think. I'm starting to learn a lot about economics, and now that I'm starting to establish reasonable credit, I'm finding myself very interested in credit modeling.
I plan on staying here for a while and building the oh-so-important industrial experience up, and maybe finishing my PhD later on in life if I can convince a company to fund it. I don't think I can ever go back to the grad-student stipend.
To the PhDs in this thread, don't you think it is wasteful to go through the PhD process to then work in industry?
Wouldn't a Masters be enough?
>To the PhDs in this thread, don't you think it is wasteful to go through the PhD process to then work in industry?
>Wouldn't a Masters be enough?
Absolutely. The reason I went for a PhD is that I thought I would go to Academia. It wasn't until I reached the end that I learned Academia is a bunch of weebs with no career prospects.
> 20
> finance and applied math
I want to get into quant because ~money~ but while I love and really throw myself into calculus/analysis etc, stats just bores me to tears. Most of my financial-maths courses so far have focused on basic stats, though, and all the math courses I enjoy seem to be aimed at either physics or pure-math students.
If any of you are in the field, what can you say about the types of math that are used, what areas should I be looking at? All the econometrics texts they assign us are just plug-and-chug shit dumbed down for economics students. Also, any programming language it would be good for me to self-teach (already have some python)? I'm looking to do a fair chunk of self-study over the summer.
tl;dr if I want to do quant, am I fucked if I don't like stats?
How do you expect to produce anything reasonable to hand to an employer if you don't like/can't do statistics? You can always try the private consulting route, but again: statistics.
More importantly, you claim to "love and really throw [your]self" into "calculus,analysis,etc.," but somehow dislike statistics. You clearly have very little understanding of both calculus and analysis because you would realize quickly that statistics is based heavily upon and employs ideas from both. I fell in love with statistics precisely because it was application of, albeit very specialized areas, of analysis. What separates statistics from, say analysis( and probability) is the type of questions you aim to answer.. Testing hypotheses, constructing confidence/credibility estimates, etc.. Why don't you try taking some of the graduate level statistics courses and see if you still feel the same way.
Whats popular these days: memes
What are you doing: working
what do you want to do: live forever, mine and explore V.R and the Universe.
Oh, and to answer your question: yes you are fucked. You will never get a job as a quantitative analyst if you can't do statistics.
At this point in the game you should also be comfortable implementing various ML models in at least one programming language..
also going to a top 5 but im a prefrosh, how much would you get paid? I'm sort of interested bc i know I like stem but don't know what to major in
graduated with a bachelors in applied math from a super shitty state university in 2012. 3.8 gpa.
currently make $16/hour doing manual labor in a factory. :^) kill me now.
>Graduated with a 2 year technical degree in programming
>Passed college alg with a low C
>Have to use a calculator to multiply any number larger then 12
>Started out as a SQL DB admin
>Currently making 55k/y as Java/ dev for startup
Why are you all doing programming? It has so little to do with math. I suck at math and i'm great at it.
There's different types of programming where more math helps. My first year I made 110k with unlimited time off and this second year I already have multiple people working under me and I'm making 130k
You are better at math than you realize.
There's more to math than arithmetic and algebra.
Makes sense. I can believe that.
>You are better at math than you realize.
To me math is just memorizing hundreds of unrelated formulas for specific instances, and I can't do memorization.
With programming you only have to learn the basic tools (loops/stacks/pointers/exc) and you can do just about anything.
No remembering ax+by=cz but only when a is not negative or when b is raised to the second power.
Bullshit, unless you're doing graph theory or statistics most math is orthogonal to the topics covered in CS.
The one exception is scientific computing but most people don't get a job doing that.
I am the first of your quotes.
>Why are you all doing programming?
As you can infer, from me not having an education in software or anything related to it, I self learned all I know about computers and programming. Why? Because I like programming.
I understand that a lot of people on Veeky Forums for some reason portray themself as one dimensional people who only care about math, but I'd like to believe that most are not like that.
I've wanted to work in the software industry before I even got into university, way before I got my degree. My education in mathematics certainly gives me more weight as an applicant first because of the qualification and second because nowadays to do software you need to know statistics, probability, calculus, linear algebra, algebra in general, and number theory to some working capacity. Unless you are doing video games or web shit or I don't know.
Anyways, short answer is because I want to and because I can.
>It has so little to do with math
It depends, let me tell you a recent story.
I was given a project that involved taking an excel spreadsheet which did certain calculations using excel functions, and turning it into a user application as to be more portable and user friendly, and more importantly, could automatically perform these calculations on different variables many times per second.
The specifics of what was going on are complex but it was like this:
A certain function did a procedure that performs 3 IF checks on the input (floating point number) and then did a multiplication. This is because this function is generalized to work for any input.
However, I noticed that in my case the input was always going to be a 4 digit number, because the input was a rounded percentage (0% - 99.9%) that would be multiplied by a 100 before being fed into the function,
Cont.
Cont.
Just by knowing this I could write a new algorithm that only needed to perform 1 check (if the number was odd or even) and then just do addition, a much lightweight process compared to multiplication.
If the original procedure took 100 miliseconds, this took at most 1. And this is significant when the entire point of turning the spreadsheet into a standalone application is to be able to perform the computation over and over again with a huge set of varying variables.
Now, I am not saying that I did some crazy number theory, but I certainly was thinking math. Actual math, the math of logic and analysis.
> I suck at math and i'm great at it.
Programming itself does not require big knowledge of math, but for one thing, you would never be able to take my job.
That is not a bad thing but you should know that you are limited in your career by how much math you know and what qualifications regarding math you have under your belt.
Thank you that's an interesting take on the subject.
Hearing you guys talking about going into programming casually sounds to me something like "Yea I just completed my PHD in physics, I can't wait to start working at McDonalds"
Doing what user?
I am a quant.
Get ready to get fucked in the ass and to lie a lot to make people feel better.
They'll pay you to rework the risk until the numbers look nice enough for their liking regardless of reality.
>They'll pay you to rework the risk until the numbers look nice enough for their liking regardless of reality.
If you hate it so much, couldn't you just go retail?
Well, they are all probably rich enough to not give a fuck if they lose a few times.
Pro-tip: look if they're employed in a math field.
>Bullshit, unless you're doing graph theory or statistics most math is orthogonal to the topics covered in CS.
>The one exception is scientific computing but most people don't get a job doing that.
>Doing what user?
"Data Science" which can mean a lot of things but what I mean in particular is "Applying Machine Learning and Natural Language Processing at various levels of our business to automate many tasks, quality control of our data, and adding features to our products".
What you have described is not math at all. Not even a little bit. Math is solving logical problems under a given set of parameters and formulating arguments.
> (You)
>Bullshit, unless you're doing graph theory or statistics most math is orthogonal to the topics covered in CS.
>The one exception is scientific computing but most people don't get a job doing that.
I do Data Science which is mostly Machine Learning applied to business problems. My particular work has little in common with CS except that we both use computers.For my job calculus, linear algebra, basic statistics are very useful. I must admit I've started studying more discrete math, logic, formal language theory, finite automata, and algorithms on my own though just because it helps me see the connection between the high level ideas and the computer code I need to write to prototype those ideas better.
How? Who has the data?
I graduate next year with a bach psych and minor math.
What can I (not) do?
I don't even enjoy the depths of both fields. For psych I took all the cognition/memory classes I could. For math I took 5 "problem solving" courses. Like baby proofs and discrete maths. Calculus was a fucking drag.
American Mathematical Society, check out unemployment numbers and other stats here: