In the News
How the Circle Line rogue train was caught with data
Great data detective story! For months, a train line suffered from mysterious disruptions and created confusion and distress. Here's how a team of data scientists saved the day.
The secret to smarter fresh-food replenishment? Machine learning.
Fresh food accounts for up to 40 percent of grocery store revenue. It's also perishable, demand is highly variable, and lead times are often uncertain. This McKinsey article explores machine learning approaches for an industry that isn't known for cutting-edge data science.
Economists are prone to fads, and the latest is machine learning
Is it really a useful tool or is this latest craze distorting economics?

Sponsored Link

Level-up your Python workflow with Mode
Mode is a SQL editor, Python notebook, and visualization builder all rolled into one. Explore data with SQL and pass results instantly into a Python notebook for further exploration and visualization. Pick and choose output cells to present to others, or send the whole notebook—you can even share with people who don't have a Python environment set up.

Tools and Techniques

An Interactive Tutorial on Numerical Optimization
This interactive tutorial demonstrates some basic numerical optimization algorithms. Includes useful descriptions, linked references, and a Github repo. Highly recommended.
Reproducible research: Stripe’s approach to data science
Here's how Stripe has operationalized data science, including the thinking behind its decisions.
Probabilistic Programming
Introduction to probabilistic programming, starting with the basics: what it is, what it's good for, and how to use it. Includes useful references and applications. This is very well done.

Resources
What makes Bach sound like Bach?
MusicNet is a collection of 330 freely licensed classical music recordings with curated fine-level annotations. This new dataset is designed to teach classical music to algorithms.

Data Viz

A Deep Dive into Geospatial Analysis
Many datasets have some kind of geospatial component to them. Python provides a rich toolset for working in this domain, and recent advances have greatly simplified and consolidated them. This is a great tutorial that explores a dataset of AirBnB locations.
Highlights from IEEE VIS’16
Enrico Bertini from the Data Stories podcast collected some of the best links, notes, and projects from this year's IEEE VIS conference. This isn't just a list. Each reference includes a useful synopsis.