Did you try to get into Data Science and have no luck?

How good were you?
>Do you know all the math notation?
>Can you implement any paper in Numpy?
>Do you have high Kaggle scores and repos of interesting projects?

All of this or more and they still didn't look at you because you're a White male?

Attached: DbOm4oiUQAAq52V.jpg (1200x739, 203K)

Other urls found in this thread:

paperswithcode.com/
twitter.com/SFWRedditImages

>Data (((Science)))

Anything that has a science at the end of it is a Mickey Mouse degree

I know it's a meme and Jow Forums rightly shits on it.
I mean more like getting a job building different models. I think it would be pretty cool job, dude.

No, I didn't try, even though my background is in pure mathematics. """Data scientists""" are either actually doing statistics, or they're not—and if they're not, then they're dangerous. I used to work for a consultancy that had some """data scientists""" who were using genetic algorithms to p-hack without understanding how fucking stupid this is. I quit the company within a few weeks.

90% of data science is just retards plugging CSV into hacked up Python scripts until the plots look like what they like.

>p-hacking
>plug CSV into hacked up Python

I'll take $120k a year doing either

>I'll take $120k a year doing either
You say that, but when you're sat in a presentation and your coworkers—none of whom have statistical training—are reporting their hacked-up statistics to credulous clients, you'll die inside. Or, at least, you should die inside.

I've made some good money (.fit)ting my way around.

>I've made some good money

You have a degree?

>Die inside
How bout you? You got the job because of how much math education?

I have only a high school diploma. I can't get hired even doing fucking web development.

I am a researcher at my university working in computer vision. As such we work with models and a low to medium level of maths. Reinforcement learning (and physical applications to the robots we hsve in the lab) is also cool. To be honest, there are interesting problems in this area but I feel data science rn its similar to the .com crash in the 90s to 2000s where investors jumped on the fomo boat, but the networking ideas of the systems are neat. Corporate hype is bad (i.e shit like viz or phacking like others said) but the underlying math can be interesting. Read papers and read their code on github rather than focusing too much on math notation or trying to implement every concept you read. After reading enough cose you'll learn some cool tricks and learn other people have saved you some pain.

>Read papers and read their code on github

paperswithcode.com/

>how much math education?
MMath (Oxon.), incomplete Ph.D. (Cantab.), though I'm trying to finish it in my spare time. You'll likely struggle to get hired without a degree these days (especially in an area like """data science""").

I was hired because of my background and because I was supposed to help them develop their statistical models. It turned out that they had no interest in giving me the resources needed for this (e.g., blocks of uninterrupted time not spent in a loud open office). I also suspected that they were only interested in Bayesian methods because it would make it easier for them to lie. I quit and became a programmer. Be wary about what companies in the area say about themselves; like AI, a lot of companies will just bullshit about it.

deep learning (neural networks) are overused, most people in ML are just churning data through neural networks and are incapable of coding any ML algorithms by hand

Thanks user :) i love that website, should have added it. arxiv-sanity.com is good to sift through arxiv or find stuff, use scihub for more obscure shit. Some papers have code but arent in paperswithcode unfortunately.

>You'll likely struggle to get hired without a degree
> AI, a lot of companies will just bullshit about it.

Sometimes I want a job doing it. Sometimes I think I'd rather Freelance web develop or work in a restaurant or Walgreens part time and have human contact and learn and build projects doing exactly what I want to do on my own time.

> UK degree
I immediately got the impression you didn't know anything, now I know for sure

that sounds like fun until you have to be outside at work for 9 hours (including lunch), then spend another 2 on commute, 1 on shower/prep etc and then you're beat by the time you're home to work on your stuff

stay salty

Attached: ebin.gif (500x364, 983K)

Why quit the PhD friend?

Why is data sci a meme? I hear the huge salaries and feel jealous that all they do is fuck around in snek language

Literally a field about bullshit. Want to do what their marketing claims and more? Then study real math, like statistics, and gain experience and actual knowledge in whatever scientific field you want to specialize in.

Data scientists literally would not exist if it weren't for marketing retardation.

I wanted to try an build a neural network or something which could play an RTS like CoH, but I wasn't motivated enough

How would you plan it?

that's what it's about though retard
it's not exactly their fault that everyone has access to mathematical functions they don't need to fully understand.

>Why is data sci a meme?

Though neural nets are very real and amazing it feels like 'AI' and 'data science' are in a hype phase. Probably because we are near the end of a long credit cycle.
When I read articles about it, and even instructional videos, it's hard to sense the content of what is actually being discussed. An indicator there is confusion and hot air.

No one owes you a job. If you are that good then companies should be lining up wanting to use your products for big $$$. If you just say you are good and then beg for a job you are below code monkey pajeet

I forgot to put in the OP, I was wondering if it's possible to get a job without a degree if you're a White guy.

If data science is a meme then DSP is too right?

Attached: dsp.system.gif (542x185, 3K)

It's not a meme, don't ever listeb to Jow Forums ever, newfriend.

The term "data science" is a meme, and the majority of works in this area. Anyone that tells you is a meme or he is a neet or he is a "self taught" data scientist. The field is almost nothing new in theory, but it is a multidisciplinary field taking from almost any mathy field (statistics, linear algebra, control theory, logic and most of all information theory) and they rebecome hot thaks to modern processig power. Even kalman filtering or kolmogorov-wiener prediction approach are tools of data scientist even if they born in control theory and system identification

yes, it is possible, now fuck off

>fuck off

Why are you mad? Rage is a stressor. Stress is a killer.

I'm pursuing ML as a hobby and wonder how hard I should work towards making myself appealing to employers.
You've seen anyone get hired in DataScience without a degree recently?

Attached: machine_learning.png (371x439, 28K)

It depends on your background. You can learn ML directly and being kinda proficient but you'll be the pajeet of the field. Just like it doesn't make much sense learning statistics without learning calculus and linear algebra. Sure you can do it but no one should hire you on the same level no one in their right mind would hire a self taught statistician

That's a joke, you know you must not take it seriously?

I'm glad they aren't letting white males into tech. I work in tech and I'm not trying to get shot up

Attached: KSxkh1C.jpg (960x720, 69K)

Data Science is just fancy name for Computer science with a focus on statistic.
If you are good on probability and statistic, data science is a piece of cake.

Sure, like CS is just applied math, you retard

Weak link token hire

big brain AI researcher here