— In the News —
It was a sad week for humans. Google's AlphaGo defeated the world's best human GO champion, 2-0. The New York Times covered the first game with some thoughts about AlphaGo's performance as well as the politics behind the match. It's worth noting that, so far at least, the politics are all human constructs.
Meanwhile, Garry Kasparov, one of the greatest chess players of all time, was recently interviewed about AI and how things have evolved since his defeat to Deep Blue in 1997. This is a very smart and insightful interview.
Data is essential at Airbnb. They call it "the voice of their users at scale" and Airbnb is on a mission to make sure everyone in the organization can tune into that voice. This post from Airbnb's engineering blog explores the features of the Data Portal that will help make that happen. For anyone that wants to be part of a data-driven organization, this is a must-read.
— Sponsored Link —
ActivePython comes pre-bundled with essential tools for data preparation, analysis, visualization and machine learning. Connect to your data sources and start being productive with all the hard-to-build libraries ready to go. Integrate your analysis into web apps, test for production, and train complex neural networks, all with one Python solution.
— Tools and Techniques —
Here's a step by step guide to building your very own hotdog / not-hotdog classifier as featured in a recent episode of Silicon Valley. Crazy as it may be, someone also created an app for this that quickly went viral. Have fun with it!
A story of pi, prime numbers, and complex numbers, and how number theory braids them together. This is a fantastic half hour video.
Great article about the development of a production Machine Learning system. Starts with a discussion of the problems being solved and then explores the thinking behind a variety of experiments that led to the final implementation. This is a very worthwhile longread.
— Deep Learning —
It's easier than ever to get started with deep learning but there are some key techniques for making it actually work in an application. Here's a collection of tips and tricks for applying deep learning to real-world problems.
— Data Viz —
From Matplotlib to Seaborn to Bokeh to Plotly, Python has a range of mature visualization tools to choose from. In this talk, Jake VanderPlas offers an overview of the options to help you choose which to use for any given visualization task. As you'll see, there are a lot of options and Jake does a great job of organizing and presenting them.
— Conferences —
If you're interested in AI and can be in London on June 20-21, you'll definitely want to check out the lineup of speakers, panels and workshops at CogX London. This looks like a great conference.