ISSUE 305 ·   September 29, 2020        

Insight

The Hardware Lottery

This is a great essay that keeps showing up in my Inbox. It explores how the success of an idea is dependent on the software and hardware that's available. That's especially important for the evolving collaborations between the ML, hardware, and software research communities.
Sara Hooker

Sponsored Link

The notebook you’ll love to use

Deepnote is a new kind of data science notebook. Jupyter-compatible with real-time collaboration and easy deployment. Oh, and it's free.  

Tutorials, Projects & Opinions

Unpacking The Data Hype

Sarah Nöckel, an investment manager at Northzone, explores the latest in data tooling software. Covers data pipelines, data catalogs, data collaboration, data quality and more.
Northzone | Sarah Nöckel

RecSys 2020 - Takeaways and Notable Papers

Eugene Yan's write-up of the recent RecSys conference is a great introduction to some of the latest thinking about recommender systems. This is a well organized post that covers a lot of ground and includes links to worthwhile resources along the way.
Eugene Yan

Introducing TensorFlow Recommenders

TensorFlow Recommenders is an open-source package that makes building, evaluating, and serving recommender models easy. It helps with the entire workflow of building a recommender system and aims to do that while being easy to work with and learn.
TensorFlow Blog

What makes a good estimator?

Gentle introduction to modern effect estimation.
MultiThreaded

Productive Analytics with Data Quality Governance

Nubank has seen tremendous growth since launching 7 years ago. With more than 25 million customers, it's considered to be one of the largest fintechs in the world. This post offers an inside look into how Nubank manages data quality using a governance framework.
Nubank | Ariane Hoffenberg

Code & Tools

• Microprediction - Tap into the collective intelligence of community contributed time series algorithms, or add to the intelligence.

• Hivemind - Python library to train large neural networks across the internet. For instance, train one huge transformer on thousands of computers from universities, companies, and volunteers.

•  modelstore - new Python library for versioning, exporting, and storing machine learning models.

•  ipygany - a Project Jupyter widget for Scientific Visualization and 3D data analysis.

Career

6 Red flags from doing 60+ technical interviews

Interviewing? Here are some key things to watch out for.
interviewing io blog

Resources

Scientific Computing in Python

Great introduction to NumPy and Matplotlib by Sebastian Raschka. Includes 10 video tutorials and detailed notes for each.
Sebastian Raschka

And Finally...

What if all the U.S. covid‑19 deaths had happened in your neighborhood?

The U.S. recently passed 200,000 confirmed COVID-19 deaths. To help put that in perspective, this data visualization shows what that would mean if all those deaths had happened near you. If you don't live in the U.S., enter a city you may be familiar with (e.g. "Seattle, WA").
Washington Post

Data Elixir is curated and maintained by Lon Riesberg. If you need help on a data project or have a suggestion for the newsletter, reply back to this email or grab a spot on my calendar >>

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