I'm pretty new to data science and want to get my hands dirty. What are some good publicly available data sets to play with?
I found hundreds of links to public data sets here:
https://www.quora.com/Where-can-I-find-large-datasets-open-to-the-public
For health care claims, there's the United States CMS 2008-2010 Data Entrepreneurs’ Synthetic Public Use File (DE-SynPUF)
There is a public chess set at http://www.top-5000.nl/pgn.htm. You can use it predict the relative piece values in chess and more.
There are multiple datasets available in the UCI machine learning Repository. You can find this here
It depends on your interest. As others already suggested www.kaggle.com is a good place to start if you want to solve well defined problems. There is also a good community there You can learn from them. But data here is mostly clean which is not the case usually in real life. I would also suggest http://www.drivendata.org/ and https://www.crowdanalytix.com/ if You like competitions.
If You just want to play around with your new skills find any data source about the topic You are interested in (like https://www.quandl.com/ for financial and economic data) and apply what you learn for any data there.
as another user already said, this is entirely relative; on that note, it is still fun, and also a great way to learn about new datasets, etc. below are some of the more ridiculous datasets that i've come across, including my personal favorite, rat mapping in nyc!:
Rat Information Portal:
http://www.nyc.gov/html/doh/html/environmental/disclaimer.shtml
Percentage of Adults 65+ Who Have Had All of Their Natural Teeth Extracted (by State):
http://kff.org/other/state-indicator/percent-who-had-all-teeth-extracted/
UsingR
ordatasets
, but maybe you need to investigate some topics of your interest: Economy, Stock market, wheather, public data, medical data... – Enrique Pérez Herrero Sep 1 '15 at 18:47