Is data science the career of future?
What does one have to do to become a data scientist?
>>54560929
learn data science to become data scientist
It's a meme
>>54561233
That's true, but what should my masters be in? Computer science? Statistics? Economy?
>>54560929
> statistician
>>54561275
yes, this, also see this>>54561252
Sounds like a meme
Data is just data, how the fuck is it a science now?
Wow, so many jealousfags in this thread.
>>54561333
Yup, totally, go ahead and study erm.... data science! Sounds like a great field after your in-depth explanation of what it's all about and why you're so interested in it!
I'll stick to algorithm science, never was much of a data guy, but now we can collaborate! You give me the data, and I'll input them into my algorithm, then we'll give it to the results scientist and he can give it to media scientists
fucking buzzword swallowing cuckold, you were never made for this field
sage
>>54561333
I would expect these salaries to plummet once employers find out the meme
>>54561413
my school started to offer the degree. Its essentially cs but with an intro machine learning and R class with some "information science" stuff mixed in
> buzzwords
>>54561302
Chemicals are just chemicals, how the fuck is it a science now?
>>54561333
I fucking hate these lists. They ruined every single medical profession
Learn a lot of analysis (calculus) and statistics.
Go on Harvard's website and see what set of courses students take for Data Science or MIT or another school. http://cs109.github.io/2015/pages/syllabus.html
>>54561333
Data Science doesn't pay that much. Most of the jobs are for government and they pay nothing, like $80k/year if lucky. In finance you need a PhD or else they won't even look at you for Data Science.
Elastic Search is paying support engineers $160k/year to start. For support. Fuck Data science unless you really really love analyzing data.
Also fuck salaries completely. The best is CTO/CISO on the board where you get gigantic bonuses and then all the PhD Data Scientists and Developers underneath you get paltry wages.
>>54560929
It's a marketing term for statisticians that can program
>>54561934
> implying statisticians can't program.
A data scientist is a data analyst that doesn't have routine work assignments.
To get a job as one, you need to talk to the directors like it's alchemy, and obfuscate exactly what you're doing behind cutesy phrases like 'dark magic' etc., and use esoteric statistics & graphical model nomenclature that the business managers didn't hear about in their Math 330 Probability & Statistics course so they will believe you are inventing AI that will increase profitability 300% for them.
There are now universities with data science programs. Some companies have internal data science training as well.
I have deep suspicion around the bulk of it (least in the current state) is geared towards IT snd business analyst that are using the data for business intelligence insights. Translation... a bunch of middle aged business analyst that bought into the buzzwords learning surface level statistics/heuristics that now think they are a 'data scientist' and experts in statistics, and predictive analytics.
I think all of (what I described) is trash and isn't 'data science' or even statistics. It's crankry with lots of handwaving. Non-rigorous and people with such shallow understanding shouldn't be marketing themselves as data science experts.
>>54561908
>Elastic Search is paying support engineers $160k/year to start. For support. Fuck Data science unless you really really love analyzing data
Wow, that job title, are there engineers for everything now?
>>54562744
and yet you're only a board posting engineer
what a meme profession
>>54560929
well "data" usually means the exhaust of internet traffic so you'll probably start out in a virtual septic tank (marketing)
>>54564164
meme paychecks in perpetuity are fine by me.
>>54561413
This here.
This thread is full of shills, I've been seeing these data science shill threads spawning these days with clearly propagandistic pictures to sell the 'why data science best science' meme, where the term remains undefined and it just remotely rings a bell as some application of statistics with a little bit of programming. And a bunch of replies with no god damn information whatsoever "yes data science good" "I like data science, data science pay money" "data science niche science".
Beware of the shill and the buzzword fellow /g/ents
>>54561259
Economy if you want to be quant, actually they would much better prefer mathematics
Computer science if you want to work for non-economic fields.
>>54562009
This guy gets it.
But remember to throw in a soothing familiar word from elementary statistics every now and then so they can nod when they hear a vaguely familiar word and feel intelligent and flatter themselves into hiring you.
>>54564324
What the hell are you trying to imply, that this field is a conspiracy theory?
>>54564738
Didn't you know Elizabeth II is a reptile?
>>54564738
Nah, I'm just saying that somebody's making propaganda for a field that seems to need people to even survive (and which is probably a meme buzzword field), as can be seen by the pictures that always come along these threads and the absolute lack of relevant information as to what the field is about.
>>54561302
>Data is just data, how the fuck is it a science now?
my sides
best post of the day
>>54562009
that's literally any high paying career
>>54565293
>somebody is making propaganda for a field
>on 4chan
You seriously overestimate the importance of the shithole you are posting on, senpai...
>>54561333
That's inflated.
DS is in a better shape than the stale CS meme
It's a niche of CS that isn't filled with codemonkeys and soccer moms so there is actual hope of making more than 40K/year if you can do more than spout buzzwords.
If you're not focusing on a highly specialized and marketable niche of CS (with the exception of those graduating from top institutions) you're doing it wrong.
Data storage keeps getting cheaper and retention keeps growing. Analyzing large amounts of data can point to micro/macro trends that were not accessible in the past. It's only natural that a profession grows out of this trend in the market.
Ofcourse, there are lots of webdev hipsters regurgitating buzzwords like datascience but there is a high paying career behind the general concept. Data Science refers to a range of careers so it has become a buzzword much like CS
well, im working in data science (not really data science, it's more like research) because programming centric jobs (ie. Software Developer, Software Engineer, Programmer) in industry is soul suckingly boring and statistic centric jobs are boring as well. thankfully programming is still a requirement so i get a nice balance of programming and statistics.
one thing that I really like about my job is that I get to work on prototypical stuff. there's a GOOD chance that this shit isn't even going to matter in 5-10 years, and that 2% of the world's population isn't going to even know about this, but there's also a chance that my team can be "on to something".
>>54566177
Can you tell us more about your job? Also, what aspects of it you find fun?
>>54564337
>>54561259
I study physics, and I have a part-time job in data science on the side already
BIG DATA! xDDDD
>>54567746
i work in the states in a fraud detection team. in short: we have a list of transactions/"activities"/summaries, and it's our job to write an algorithm that detects fraudulent activity.
i find knowledge discovery and data mining a lot of fun. you have an abundance of data and you essentially just sift through it all and find interesting (and small) nuggets of information. apply simple statistical verification methods (sometimes i write them, sometimes i just feed the data through some gui like weka or rapidminer if its a monday cause fuck mondays) to make sure that your findings are sound, and then report that on your next demo just to lighten up the mood before you actually demo what you've been working on. these small nuggets are usually written to a backlog for investigation later. i dont really spend a lot of time on this. i usually just do it on mondays, the beginning of the sprint, or near the end of the sprint or when im feeling a little bored of my current project.
i find projects involving classification and cluster analysis a lot of fun as well. they're a lot more time consuming than basket analysis and data mining, but a lot more rewarding as well.
visualization isn't fun at all, so i usually just give the intern the visualization projects. depends on the intern though.
i hope i answered your questions.
>>54569300
What kind of algorithms and tools are good for that kind of data?
>>54569408
you should never look into algorithms and tools based on data. a lot of data analysis and "data science" is mangling with data, cleaning it up, transforming it into something that is analyzable.
context matters as well.
most of the time domain knowledge and a basic understanding of statistics beats all kinds of tools. knowing what frequent pattern mining, clustering, classification, and outlier detection are would be very useful as well.
there's no 1hko for data. a lot of libraries and gui's do small things and they usually only accept one or two formats so its up to you to make sure they work. with that said, weka is really user-friendly, and so is rapidminer. don't rely on gui's though.
i recommend the book "data mining: concepts and techniques". it is a very good survey book for frequent pattern mining, data warehousing, clustering, classification, and outlier detection.
this subject seems like it would take years to fully make use of it on a practical level
>>54571150
stfu
>>54571150
Sorry, not him, but I just started a data analyst job doing fraud detection at a bank. Was hoping you could share some tips about the job.
