— In the News —
From tricking search engines to injecting bias into training data, data manipulation isn't new. But with data increasingly at the center of every decision, the effects of manipulation are more and more concerning. Here's a look at the issues and how we might build technical "antibodies."
A hot new startup intends to use AI to keep you healthy. That makes a lot of sense until you start thinking about the details. Would you pay to have AI continuously monitor your health?
— Tools and Techniques —
There are important insights here about the need for data availability in a computing world that's increasingly dominated by probabilistic inference. This is a thoughtful longread that includes ideas about data fiefdoms, exchanges, brokerages and cooperatives.
David Robinson at Stack Overflow recently showed how Python is the fastest-growing programming language, largely because of the growth of data science. In this post, he explores how R has grown, including its growth across industries and the specific packages that are commonly used.
Len Kiefer is the Deputy Chief Economist at Freddie Mac and his blog posts are fantastic explorations of the economy, housing and mortgage markets. In this post, he shows how to aggregate several million loan level records into useful summary graphics to provide insights into the U.S. mortgage market. It's a "data wrangling and visualization extravaganza."
Jupyter Notebooks are useful but they're not perfect. Here are a few issues to think about with suggestions for how to work around them.
— Deep Learning —
Grant Sanderson's Animated Math blog is a super popular place to learn about complex topics. In his latest video, Grant offers an introduction to neural nets and it's one of the clearest introductions you'll find anywhere.
Here are six key developments in the history of deep learning with background about the breakthroughs, the people involved, and code snippets to bring them to life.
— Data Viz —
This new site showcases a variety of Python chart types and includes use cases, examples, variations, and lots of sample code using Seaborn and Matplotlib. This is definitely worth bookmarking.
— In Case You Missed It —
Be sure to catch the most popular links from last week's issue...