Data Scientist. Thoughts?
Data Scientist. Thoughts?
why does this fucker look so smug
Looks like a real industry cucked beta onions boy faggot.
Interactions engineer is superior
maybe if you're an onion rings boy
>have a degree in math or statistics
>Jr positions require MSc or PhD
>half the posted data science jobs are actually consultant development jobs.
This was my experience.
well he has your data
God I hate marketing fags.
very small pp
I'm sure this will be a great thread.
It means they have no internal systems or processes so you need to build them yourself in addition to all the data science stuff you do.
Currently doing a PhD in text mining for finance. Not planning on postdoc. All jobs remotely machine learning related are data science. Will have to work on my visualization and stats skills.
oh shit, a meme I relate to. If the person was a tranny that'd be me. I like my job though
The same I think about code artisans and language evangelists. Faggotrons trying to be special.
>first time with laravel
Who the fuck comes up with these things. "Code artisan" is something I would almost use to mock a developer.
it means that they aren't going to give you a reasonable timespan or any documentation for their existing codebase so they want you to hack it together
Can you talk a bit about specifically what you're doing/how text mining can be used in finance? Truthfully I've always thought that full text analysis/sentiment analysis/whatever is still way too immature to be used in any capacity outside of research, but truthfully I don't know much about the state of data science/ML today.
>Data Scientist, the sexiest job of 21th...
stopped reading right there
I don't get it....
Not him, but aggregation and analysis of market attitudes based on things like news articles and blog posts is a thing
>you sit in an office writing code to comb through 100MB CSV files for 8 hours a day
nothing turns girls on more than a lively debate about p-values
also gotta love this one
>influence without authority
translation: your manager probably knows nothing about stats so be prepared to eat shit in meetings for months before anyone understands what you're doing
I'm a statistics and data science post doc. The job market is as big a meme as this infographic. On the upside about 90% of the work is remote so at least I can avoid staying in LA or NYC for the rest of my life.
Looks like the latest basedboy field
Just say "machine learning" and every fucker will bow to you.
A bachelor's in electrical engineering still requires more knowledge and higher IQ than getting a PhD with a focus on machine learning
looks like he is dressed up to take his wife's son to ballet class or interview his wife's newest tinder date or some shit lmao.
>they even type with a lisp
And he still makes less than me being a codemonkey writing java
He earns $800k/yr
His months worth of s.oylent came in the mail
so much fucking this
t. data scientist
data science was a mistake. it looks good and shiny till you are actually doing it at a company with their shitty data, non existent data infrastructure and retarded management. you have to suck up shit 95% of the time just to have that 5% enjoyment that you became data scientist for
I only gave it a quick glance, but I didn't see AGILE anywhere, so this is laughably out of date.
Is it true that there are basically two tiers of data science? The actual data scientists at big tech companies analyzing user data in the TB/PB range and doing advanced math and distributed processing, and the """data scientists""" at random companies where they do basic scans of 3MB CSVs with python libs who were hired so the upper management can jerk off about how nextgen they are at their next tee time?
>tfw you meant to say onions but onions came out
>tfw so much of a newfag he didn't know about wordfilters and will probably get triggered to fuck by discovering them
hiirooo why do you side with the onion farmers it's not fair onions make me cry
I wish people would stop trying to push this data science = sexy meme. I actually enjoy this autism work and don't want anyone joining the ranks who isn't cut out for it.
>Being a literal corporate drone its "the sexiest job of 21th century"
Tier two data scientist reporting in. My company's database is in such shambles I'm more of a data janitor.
My guess is Tata, Infosys, TechMahindra and a few of the others realize this is going to be the next big thing for them, so they started setting up body shops, getting the paper mill credentials, and writing glowing resumes for guys who just last year were doing tech support scams. When you see a job description with ridiculous requirements like 10 years experience in a technology that's only existed for 5, and a Phd for an entry-level position, that's usually what's behind it, and it's totally guaranteed to be the case if they are looking for consultants, contractors, or corp-to-corp workers.
Not a data scientist, but isn't that what Hadoop and some other tools are for?
Hadoop is, surprisingly, not a magic box that magics its way into magically making sense of your "big data" for you. Probably 90% of the things most companies use "data science" tools like that for would be significantly better off in the hands of a small team of people with a great business intelligence background and one statistics guy. That should be the starting point before you go off and even start thinking you need a "data scientist" but no, shit infographics like this cloud the order that things need to come in, and people throw "data scientist"s with tools that aren't right for the job at things that are essentially data organization tasks.
Unfortunately though, BI is even less sexy than data science and is largely filled with shit-tier, outsourced, button-clickers who don't know the first damn thing about how their tools actually work under the GUI.
Marketing for data stuff is somehow always the absolute fucking worst.
The most accurate part of this depiction of a data scientist in this picture is that the things below the waist don't matter at all
>geologist doing a phd
>work with lots of well logs and time series
>be Jow Forumstard
>suddenly I work with hdf files
>lol, data scientist
>add data science to my linkedin
>suddenly, great job offers
but I can only type "how do I do X with Y in pandas" into stack overflow and then implement what I find there.
It's an abstract kind of feel
Text mining predoc here. Noticed that too: job listing with the title data scientist can go from statistics to machine learning to visualization. The term is a meme.
And than you basically gave tier 3: the ones making graphs and dashboards that are more like front end web devs.
So far I identified 4 archetypes who are referred to as "Data Scientist":
Data Scientist (the real one)
Machien Learning Eng. (somewhere between the Data Scientist and Data Eng.)
If a company does not know which one they need, it is an instant red flag.
It's a rebranded code monkey position.
If you've met those who bear the title you know they're just barely passing for what they do and what they do isn't much.
I've met one who asked me to explain the difference between median and average. And I did, it wasn't a joke. He seemed to understand.
Most of them weren't elementary school level but they're pretty far down there.
What are you doing now?
My Economics degree is about the same thing and I’m lost with the requirements.
Most start up "Boss" are smug and full of themselves. This is a personal advice + $0.00 tip.
Yes, that is exactly true.
They even will exist within the same company.
On the upper tier you have PhD level people who write research papers and prototype algorithms, and the lower tier you have Pajeet you cleans data and worries about databases and data pipelines and shit yet thinks he's a fucking mathematician.
The meme with data science is that there is basically nothing between those tiers.
Tell me how to become one
jesus christ why dont you throw in being a total fag in there too, no seriously what the fuck is hive and pig?
Tell me how to become one
>low effort infographic with shit-colored background and a poorly drawn, equally low-effort "mascot" taking center stage
>describes a meme job, with bullet points copy-pasted from a job posting for a data scientist but with generic shit that applies to every job in the world thrown in to make it appear more informative on surface level
>depicting those who do it as generic, flavorless söyboys
>using buzzwords and meaningless phrases like "passionate" or "curious" and encouraging the use of such as recruiting criteria even though they can't be quantified or measured in any objective way
>the company who made the infographic has a name related to craft or artisanship and a "designer" old-style logo
>their description is just as full of söy memes and buzzwords as the infographic they produced
it's a doozy
My gf did a physics degree and got a job working in the city working on massive finance DBs. Basically has free rein within her team to pick and choose projects that they find interesting, and can go to IT and ask them to bump up the computing power so they can run and develop algorithms on massive datasets, and cos it's the in thing they basically get money thrown at them. Sounds pretty comfy desu
I recommend statistics knowledge and R programming skills as an adjunct to a non-quantitative major. It's what I did and now I'm the only X that can do Y, it's pretty sweet for fucking around in R until it computes.
Sounds like the average computer science graduate desu
nice graphic there friend however the real data scientist job requires literally two skills
>python OR R
he's getting paid 120k to write python scripts and SQL queries from his macbook pro in a shitty jupyter notebook with no version control
you see, there's so much technology implementation out there, frameworks for everything, a million way to do [task], that every business taking itself seriously opens up the whole internet to the people to look it up.
For a reason.
What you're doing is the basic way anybody in this new world learns a skill anymore. Universities are outdated and at home there's usually little reason to have fat databases worth analyzing, so everybody is aware this shit will happen on the job. Who cares if you know stuff by heart if you can just as well read the requirement, do a quick google and do a similar job to a professional in +1 or 2 days if the quality is the same?
Looks like a prick.
The absolute state
This is correct, with the occasional retard doctor asking for some basic spss analysis
t. Statistician that uses R
Everything that has a science in its name is not real science.
>studied political science for bachelor
>plan to study more specialised aspect of political research for masters
I can use SPSS well enough, and am getting to grips with SQL and R fairly quickly. Python eludes me for some reason and I don't enjoy learning it nearly as much as R or SQL. How screwed am I?
For doing stats R is superior to Python so you should be fine.
Is there a realistic expectation of being able to use both? Say I set myself the target of being at least proficient in using R, SQL, and SPSS where necessary. Would there be any reason/need to know Python as well, or is it just a nice extra to have? Also, is it worth learning to use Python as a general programming language, or just learning how to use numpy, pandas etc.?
Any recommended book on statistics for data analysis or other learning materials? Im an electrical engineer that looks at thousands of data points on millions of units and all the theoretical classes I took in school did not help one bit.
Infographic by Mike Tyson
LMAAAAAAAAAAOOOOOOOOOOOOOOO NIGGA WHO YOU JOKIN YOU JUST HAND OVER PEOPLES PRIVATE INFO TO THE JEWS AND CALL IT A DAY, ROFL.
>Also, is it worth learning to use Python as a general programming language, or just learning how to use numpy, pandas etc.?
That's why I choose to use pyhton, as opposed to R. Maybe at some point, I need it for something else.
I kinda did the reverse of you:
Learn python (though "learn" just means I can google my way to make shit work, I can't write a single LOC without an internet connection) and then start using R, because there's some things in my field that are only published as R modules.
Took me 2 days of googling to get the R thing working in my python scripts via r2py…
Now I don't know if I should feel clever or dirty.
It's not even fucking science. Actual engineering of literally any kind is more science than this shit.
They're all memes. I did my Undergraduate degree in Stats, Masters in Computer Science, and PhD in Stats. I'm in a Data Science postdoc now. Every legit job is DS related. Analysis is done by BS holders. There are basically huge gaps between who does the high level design of these projects and who does the grunt work. You will need a postdoc level of experience to get the former and people with a BS who think they will progress internally are wrong. Corporate is not smart enough about this tech to train anyone and there is huge overlap between the academy and fortune 100 DS ventures.
For analysts, yes. For the project designers, no. The former is a 50k/yr glorified data janitor job. The latter pays 4x that to start and peaks close to half a mil on average. Good DS consultants right now can pull seven figures in LA. I haven't been to NYC since 2013, but I would bet it's even higher there.
Should I feel bad that I can hardly write anything without an internet connection?
I started to study data science in 2007.
We WERE 15 in my PROM.
In the second biggest city in my country.
Now I teach there and there is hundred of retards, kek. Feel good to be avant-garde.
>What are you doing now?
taking time off, working on my hobbies lol
idk, the job market is absolutely fucked atm, imo. On the one hand everybody is losing their shit that they can't find enough people for the job, on the other hand they're wanking us around with those stupid ass job proposals, everybody is losing people left and right because of broken promises, and HR is rejecting applicants outright because they don't know what they're looking at.
i think the so-called IT labor shortage is self inflicted.
is this pic from a book? thanks
git gud m8
wondering the same thing, we never learned the relationships between the distributions. is there a good book that teaches you all that stuff so you can actually understand it, instead of cooking by recipe all the time?
most probability theory books explain some relationships between the distributions.
Just got a data science job anons (analyzing network data), what should I expect?
never read single math book, never attended a single math lecture in college
that's probably the problem.
i guess college really fucks you when you're smart enough to pass classes without putting any effort in.
So.. dev ops?
Bloomberg uses a NN to find and extract info from PDF documents