— In the News —
Great overview of the Big Data ecosystem by Matt Turck of FirstMark Capital. Includes a discussion of what's hot now, where things are going, and a comprehensive map of important players. Matt starts with the bottom line, "It feels good to be a data geek in 2017."
Data is under attack. It's all too common these days for governments and businesses to alter data in order to serve their own interests. This article explores the issues and how a public blockchain could be used to maintain the integrity of important data.
Gary Marcus is an expert in learning. He spent decades studying how children learn and in 2015, he founded Geometric Intelligence which focused on building better machine learning techniques. Recently, he was acquired by Uber to run their newly founded AI labs. Even so, Gary Marcus remains skeptical of all the AI hype...
— Tools and Techniques —
Here's everything you need to know about setting up analytics at your organization. Covers tools, processes, who to hire, things not to do, etc. This is an organized and well-reasoned approach by Tristan Handy of Fishtown Analytics.
Part 3 of a popular series about exploring data with Python. This part shows how to zero in on the data you're interested in, pivot, visualize it, and examine relationships between variables.
This new post in the Google Research Blog describes an approach for collaborative machine learning that doesn't use centralized training data. Interesting applications here.
Deep Learning has broad applications for medicine. This simplified example describes traditional approaches for counting white blood cells and shows how deep learning can vastly improve accuracy while decreasing costs. Similar techniques are expected to be useful for identifying a variety of other cell types and disorders.
This tutorial uses minimax and alpha-beta pruning to build a simple chess AI. Includes code, ideas for further development, and libraries for viewing the board.
— Data Viz —
It's easy to think that data represents truth but "truth" is a tricky thing. This post by Lena Groeger explores how bias creeps in throughout a data story - starting with data collection.