— In the News —
According to a new survey from McKinsey, AI is gaining ground rapidly around the world but few organizations are able to effectively use it at scale. This summary explores ways that AI is being used in business, common issues, and where it's being used most successfully.
— Profiles —
Giorgia Lupi is an award-winning information designer, co-author of the book “Dear Data” and founder of Accurat, a data-driven design firm. In this interview, she shares some of the inspiration behind her career and tips for finding the human element in data.
— Tools and Techniques —
Bugs in your data can be just as consequential as bugs in your software but unified testing philosophies are largely lacking in the data world. In this article, Josh Temple takes a look at why testing data is hard and he offers a shortlist of considerations and strategies for testing the data in your pipeline.
Great introduction to the scikit-learn library that shows how to think through a machine learning example. Covers initial data exploration, simple data visualization, creating training/test datasets, algorithm considerations and model building.
When Help Scout set out to hire a new Senior Data Analyst, they asked applicants to answer a few short SQL questions as an initial screen. This post explores the answers and serves as a nice study guide for SQL best practices.
Here's how The New York Times built their own Market Mix Model infrastructure to quantify and understand how various influences impact sales.
Go beyond pandas, scikit-learn, and matplotlib and learn some new tricks for doing data science in Python.
Nice comparison of GPU cloud service providers, including specific things to look for, capabilities, costs, and performance.
Don’t miss your chance to join data science leaders for two days of peer-led sessions and discussions that will leave you with the keys to solving your organization’s most pressing challenges. Data Science Leaders: register with promo code ‘DATAELIXIR’ and pay only $699.
— Data Viz —
In the first in a 3-part series, James Bednar breaks down the huge range of visualization libraries that are available for Python. This is a super useful post that includes insights about the origins of each library, use-cases, and how they all relate to each other.