ISSUE 340   ·   June 15, 2021

Insight

When Graphs Are a Matter of Life and Death

Charts may seem ordinary and mundane until you stop to think about the gigantic conceptual leap it took to first imagine them. And then you can't help but be blown away by the stunning ingenuity of humanity.
The New Yorker | Hannah Fry

Don't Feed the Thought Leaders

This is a fictionalized story about software but it's easy to see how it applies more generally. Feeding a know-it-all tends to create: 1) hype cycles & technical debt, and 2) exciting conference talks 🤪🚀
Earthly | Adam Gordon Bell

Sponsored Link

Ray Summit: Scalable ML & AI for everyone

Want to learn the best way to scale? Ray Summit brings together data scientists and engineers to build scalable ML & AI using Ray, the dominant platform for distributed computing. Learn about top trends in machine learning & AI, ML in production, reinforcement learning, cloud computing & more. Register to join live or on-demand.

Tutorials, Projects & Opinions

What the Heck is a Data Mesh?!

Great introduction to data meshes, starting with the idea of "data as a product." Everything flows from there — the need for decentralization, self-serve infrastructure, federated governance, etc. 
Chris Riccomini

The Rise of the Metadata Lake

Most organizations have only just scratched the surface of what's possible with metadata. But as metadata continues to grow in volume, it's becoming increasingly important to think about how it can be used and stored more effectively. Introducing, the metadata lake...
humans of data

Patterns for Personalizing Recommendations & Search

Personalization is the process of customizing each user's experience. In his latest post, Eugene Yan explores common personalization approaches for search and recommendations and shows how they work. Covers bandits, embedding+MLP, sequences, graph, and user embeddings.
Eugene Yan

Increasing Experimentation Accuracy and Speed

Etsy's Online Experimentation Science team is a mix of statisticians and engineers that's focused on what's essentially sophisticated A/B testing. This post is a deep dive into how they use a statistical method called CUPED to quickly learn which features improve the user experience.
Etsy | Code as Craft

Linear Algebra for Machine Learning

Tai-Danae Bradley has a gift for clear explanations and teaching. In this session of Machine Learning Tech Talks, she offers a friendly introduction to linear algebra that isn't a technical deep dive but is super clear if you're just getting started.
YouTube | Tai-Danae Bradley, PhD

Democratize data & scale augmented analytics

Join this webinar panel for practical advice on how to evolve your business intelligence with augmented analytics and scale data science initiatives. You’ll learn from top industry strategists and technologists from DataRobot, DataPrime, and more, on how to integrate AI and BI, and how to augment analytics to scale predictive and prescriptive analytics as well as machine learning. Save your spot.
// sponsored

Code & Tools

Leafmap

Leafmap is a Python package for geospatial analysis and interactive mapping with Jupyter. It's built on widely used geospatial and data science packages, such as folium and ipyleaflet (for creating interactive maps), WhiteboxTools and whiteboxgui (for analyzing geospatial data), and ipywidgets (for designing interactive interfaces).
GitHub | Qiusheng Wu

Resources

Reproducible Data Science

This online text offers a hands-on introduction to open, reproducible, and ethical data analysis. Covers reproducible workflows, data wrangling, exploratory analysis, data visualization, pattern discovery, prediction & machine learning, causal inference, and network analysis.
Valentin Danchev

Sign up to get Data Elixir's  data science newsletter in your Inbox >>

Data Elixir is curated and maintained by Lon Riesberg. If you have questions or suggestions for the newsletter, just reply back to this email.