— In the News —
The business plans of the next 10,000 startups are easy to forecast: Take X and add AI. This is a big deal, and it's here now. This is an insightful article that also inspired a thought-provoking editorial by the New York Times. This is a MUST READ!
Fun facts with links to sources. Guess how much data the NSA analyzes everyday. Or how much data the world's largest single database holds. There's great Twitter material here!
Nice article by Ben Lorica at O'Reilly Media about the major forces shaping today's data world. It's a survey, not a deep dive, but there are lots of links throughout the article that guide further exploration.
This is inspirational. Launched in 2011, DataKind leads a global effort to use data science in the service of humanity. It's community-driven, with regular volunteer events throughout the world. The title link will take you to a good overview and if you're up for really being inspired, check out these short videos about some recent projects >>
With sensors practically everywhere, cities have access to vast troves of information about themselves. What they do with that information is not only interesting, this article provides insights into unmet needs. Think you know what causes traffic congestion in Los Angeles? You might be surprised what the data says.
— Tools and Techniques —
Interested in building a Neural Net? This article got a lot of attention on the Internet last week. It's an in-depth tutorial that starts with the basics and provides enough detail and sample code to get you going on your own projects.
— Resources —
Free to download, this text provides an in-depth overview of data mining. The book is based on two Stanford Computer Science courses; one is introductory and the other is a graduate-level. With course slides and access to a free MOOC, this is a great resource.
— Data Viz —
This is an interesting project that colors city streets according to their orientation. It's a bit of an art project but these maps are generated algorithmically, lending this scheme to other uses. How these were done is light in detail but described well enough to be a useful start for data mapping projects.
This is a great interactive of basketball stats by Todd Lindeman and Lazaro Gamio at the Washington Post. Whether you're a basketball fan or not, it's worth exploring their approach to presenting lots of data. Play with it first and then be sure to check out Andy Kirk's insightful assessment at Visualizing Data >>