Quote:
Originally Posted by Ativan
Thanks in advanced
It is highly sought after. But of course anything can become obsolete with time. NoCal, SoCal, Seattle, Austin, Boston, DC, and NYC are probably the best markets, but there's DS jobs to be had all over. I can't comment on OC specifically. I work from home in Ohio.
Just learn SQL. It won't take that long whatsoever. Just Google around and learn it. It will be useful to you.
How long will it take to learn? Idk. How much time do you have a week to learn it? I'd advise you to Nike it up and Just Do It. Take some data science MOOCs. Read up/practice on the required math/stats. Download R and RStudio, type "install.packages("swirl")" followed by "library(swirl)" and go from there. swirl is an R package designed to interactively teach the basics of R programming.
I'm currently working on an NBA related side project with a couple of 2p2ers. Mostly just data merging and cleaning and the like so far but about to get into the analysis itself. Feel free to join us. Happy to PM the GitHub url.
I got this job in a very roundabout way and sometimes feel like a phony:
http://www.hanselman.com/blog/ImAPhonyAreYou.aspx However, I've become better and better with time. I first encountered data analysis in my sociology grad school program. I eventually landed at a tech company as a consultant. I developed my skills very minimally, saw an internal job posting for a data scientist, and said **** it and applied because I knew it's what I wanted to do even though I didn't feel remotely ready or qualified. Got it mainly because the dude in charge has no clue. Nice guy, great manager, but still has no clue about the topic itself apart from being a skilled programmer. And he blogs on DS on a well known tech site now. Ha. But anyway, it's worked out. I probably couldn't get a DS job at Facebook, but would feel confident interviewing with say, IBM.
Feel free to PM me if you want to join the project, especially if you like basketball, or just to ask any further questions.
Quote:
Originally Posted by Larry Legend
All the data scientists I know have PhD's from MIT.
I imagine it will be very tough to get a real data science job without that level of education, and very easy if you do.
Definitely do not need a PhD. I imagine PhDs would be more common in the public sector and in a market like Boston, which I believe is the most highly educated metro.
Quote:
Originally Posted by jjshabado
It's hard because 'data scientist' is such an over used term these days. But as I mentioned in the other thread many (and Id argue most) data science jobs are more about applying data science concepts to specific use cases then they are about cutting edge research.
This is very accurate. Not only that, but I'd imagine it is far more interesting to most people to solve real problems than develop new methods of analysis.
Quote:
Originally Posted by ddubois
The feedback from the internet seems to be to 1) Get on GitHub and work on something related, 2) Get on kraggle and partake in data science contests, 3) Get a nanodegree in data science from Udacity. By the way, that list is in the order of importance, not a timeline.
#1 is good advice. I would advise against #2. Kaggle competitions often involve hardcore optimization that isn't practiced in most job settings because users have long amounts of time to fine-tune their solutions whereas this would be considered inefficient in most work contexts. I can see it being a good platform to demonstrate your worth, but it's not best for developing your skills. #3 is decent advice too. Get involved in some data science MOOCs.
Quote:
Originally Posted by NF set
"Data Science" is a wishywashy term. People who spend a weekend learning Google Analytics consider themselves Data Scientists.
1)Learn to program
and
2)Learn SAS
Data science is not well defined. It is so young, includes a broad range of possibly valuable skills, and there is a huge range in the amount of talent among those who hold the title.
However, and I saw this in other posts too, I wouldn't advise learning SAS. Stick with R and/or Python. They are free, they are better, and they are more widely used. SAS may be more widely used in the public sector, but that's probably it.