Is Data Science a meme? How do I start from the basics?
Is Data Science a meme? How do I start from the basics?
James Anderson
Other urls found in this thread:
tutorial.math.lamar.edu
datasciencecourse.org
twitter.com
Eli Hernandez
Yes it is. Study actual math and statistics as well as specialize in whatever area you want to focus on (biomedical sciences, economics, etc.). Don't fall for the interdisciplinary meme and definitely do not take any data science bootcamp shortcuts. No matter what they say you cannot replace real, domain specific knowledge and experience with vague hand-wavy garbage.
No, "marketing" is not a specialization. If you go there stay there so you don't destroy something important.
Alexander Bailey
If you don't have a firm understanding and command of elementary algebra (the algebra you were supposed to learn in high school) then I would advise not even wasting your time.
If you do then go ahead and start tackling calculus and linear algebra. You can also start learning what's taught in a basic statistics course. Next grab a discrete math book and dive in. Once you're comfortable with basic and higher order logic, proofs, and counting techniques you're ready for mathematical statistics / probability theory and then regression analysis and time series. Make sure to familiarize yourself with stochastic processes (markov chains, especially). As far as programming languages are concerned, it sort of depends on what side of data science you want to join. If you're wanting to go into data engineering (pipelining mass amounts of data) you'll want to know some of your lower level languages (I would guess). Definitely learn R or python regardless of what you want to do.
Jaxon Rodriguez
also this guy's site is a fantastic resource for brushing up on or learning some of the basic math you'll need
tutorial.math.lamar.edu
check out patrickJMT on nosetube for videos
Blake Rodriguez
For most people, yes. the problem is that even if you do it properly per
you're gonna be competing with "bootcamp" idiots who can muh-gaussian and do a bit of R and python, Maybe Hadoop/MapReduce and some SQL.
Austin Martinez
I got my first job
Associate data scientist at United Health Group
did fast.ai, Andew NG's machine and deep learning course, hinton's deep learning course and lots of kaggle. Started ~14 months ago.
William Collins
did you even read my post? i'm telling him to actually learn the math. how much deeper should he go? functional analysis and galois theory? i would argue that deep of an understanding isn't necessary
Isaiah Peterson
You will have a hard time getting an entry job because you're competing with people from both extremes (maths and stats). Not to mention that the market is experiencing bottleneck with internationals rushing too.
Asher Hughes
I know that's what you're telling him. What I'm telling him is that hiring managers can't tell the difference between someone with the background and someone that went to a boot-camp. did you even read my post?
Parker Torres
This is true. Nobody wants bootcamp trash. You can get away if got have a STEM degree but if you don't have an undergraduate degree then you're basically a drug addict in their minds
Jonathan Roberts
oh. fair enough. yes i read your post but i misunderstood it
why tf did i just dual major in biology and mathematics if these bootcamp things are legit though?
James Young
I’m a data scientist and I don’t even know what the words in your post mean.
Mason Wright
Take CMU's practical data science class
datasciencecourse.org
Note the link where you can still view the lectures even if you don't have andrewID (student login)
Jordan Ross
I took a bootcamp from that cost me about 15 grand and ive been rejected from every job ive applied to because i dont have a degree or projects. Why dont they take my bootcamp seriously?
Thomas Butler
Everybody I know that took a bootcamp was hired, granted for 1/2 the price of everybody else but still they were hired. Go on github, look up the language you learned in bootcamp and start writing features for some open source project. Repeat until you're hired
Jose White
Personally attending lambda school right now for it. Good meme if you don't have any money like me.
Carson Smith
>data scientist
If you do not know what regression analysis or linear algebra are you definitely are NOT anything that deserves the title "data" or "scientist".
The user mentioned very fundamental areas of mathematics to deal with data.
Lincoln Sanders
why does everyone want to get into data science? There are other fields in CS too, you know.
Joseph Cruz
>Data science
>CS
Andrew Evans
Data science is where all the buzz is right now because of good advertising/marketing.
Camden Torres
Data science these days translates to machine learning, which most definitely is a meme. Don't get me wrong, it has its uses, but that use isn't literally everything. Sometimes (most times) other types of statistical analysis is more suitable and will provide actually, verifiable results, as opposed to ML which for most intents and purposes functions like a black box.
Non meme jobs will still be listed as "analyst".
Ryan Walker
>How do I start from the basics?
...its fucking algebra dude
Angel Phillips
ML isn't always a black box...
Camden Roberts
Yep but it can be done on excel
Justin Roberts
Become a data engineer. It's more technical in a lot of ways and closer to core IT work.
Matthew Rogers
if this landwhale can become a dietician then you can do whatever you put your mind to opie
Christopher Price
>Data science these days translates to machine learning, which most definitely is a meme. Don't get me wrong, it has its uses, but that use isn't literally everything. Sometimes (most times) other types of statistical analysis is more suitable and will provide actually, verifiable results, as opposed to ML which for most intents and purposes functions like a black box.
>Non meme jobs will still be listed as "analyst".
The fact you even say this shows you know very little of data science.
I'm finishing my bachelors in a data science and had the honor to learn under the be. The amount of mathematics that is required to even started understanding the basic ML is absurd. Most boot-camp graduates will now how to operate the sk-learn library and no more. Most of them will just throw a neural net at a problem without even thinking, they have little to no knowledge about what optimization problems make even the most basic problems, hell I doubt the even know what an optimization problem is, let alone feature engineering, structured models vs unstructured models and more
Also calling a neural net a black box is an disservice to the complexity of those models. The way they operate is understood very well its just that it is hard to understand the results as the neural net grows in complexity (e.g. adding feedback), as opposed to weights obtained by a perceptron or a linear SVM.
Wyatt Evans
I think you misunderstood what i said. I'm saying jobs listed as "data science" are (mostly) a meme, not the field in itself, which is essentially just statistics.
Again, i am aware that ML has more to it than importing tensorflow, but a lot of "data scientists" don't.
I have plenty of respect for actual data analysis, and i do quite a bit of it myself as a physicist. But in the job market, data science is (mostly) a bullshit field.
Thomas Foster
I have a BS in mathematics. Decided against grad school because I was only getting Bs and B-s. >tfw low iq.
Question is that if I have been working as a developer for the past seven years, is there a way to transition over to data science without taking a huge pay cut?
Leo Bennett
I just got my firsr DE job. Seems really good. Get to use python, linux, and a ton of other other open source shit.
Plus the pay is really good. Its just a hard field to break into.