ISSUE 274 · March 3, 2020

In the News

Largest gift in Berkeley’s history will create a ‘hub’ for advancing data science

In a sign of the times, UC Berkeley was recently gifted 252 million dollars for the construction of a new "Data Hub." The anonymous donation is the single largest gift in Berkeley’s history. An additional 300 million dollars will be raised to complete the center. Here's what they're doing.
UC Berkeley News

Tools and Techniques

Crash Course in Game Theory for Machine Learning

Game theory is experiencing a renaissance that's been driven by the evolution of AI. In this article, Jesus Rodriguez of Invector Labs walks-through classic and new ideas that data scientists should be aware of.
TDS | Jesus Rodriguez

Improving Data Discovery for Data Science at Spotify

Spotify's internal library called "Lexikon" simplifies data discovery and helps Spotify's data teams make faster and better decisions. This is a great deep-dive into the use-cases, decisions, and iterations that drove the development of Lexikon and its success.
Spotify Labs

Aleph - a suite of data analysis tools for investigators

Aleph was developed for use by journalists in cross-border investigations. Think of it as a search engine for digging into disparate data sources. Looking for answers in a trove of PDFs, emails and Word documents? This is a good project to know about.
Aleph

A Primer in BERTology: What we know about how BERT works

This survey paper describes what is known to date about the famous BERT model, synthesizing over 40 studies. It also provides an overview of proposed modifications and directions for further research.
arXiv

Alegion Launches Alegion Flex for ML Experimentation

Alegion Flex is the first ML data labeling solution designed for data science teams in their experimentation phase. It provides the flexibility and predictability needed for teams to focus on rapid iteration, not just on production initiatives. See how Alegion supports faster experimentation, unmatched flexibility across use cases, and better control over data labeling costs compared to traditional solutions.
// sponsored

Resources

The fastai book - draft

This set of Jupyter notebooks are the basis of the upcoming book by Jeremy Howard and Sylvain Gugger: "Deep Learning for Coders with fastai and PyTorch: AI Applications Without a PhD."
GitHub | fastai

Pattern Recognition & Machine Learning algorithms

Great collection of Jupyter notebooks with clean and easy to follow Python examples for each chapter of Christopher Bishop's text, Pattern Recognition and Machine Learning.
GitHub | ctgk

Data Viz

Mapping coronavirus, responsibly

Using coronavirus data as a vehicle, this article by Kenneth Field explores common issues and best practices for visualizing map-based data. Covers projections, choropleths, symbols, and things to avoid.
ArcGIS Blog

Observable for Jupyter Users

Observable and Jupyter are both computational notebooks for doing data science and visualization but they're completely different under the hood. This tutorial walks-through the biggest differences to help Jupyter users make the most of Observable.
Observable

Conferences & Events

Strata Data & AI - San Jose, CA - Data feeds AI; AI makes sense of data. So it also made sense to combine the O’Reilly Strata Data and AI Conferences—covering two of the most pressing technological trends of the decade—and giving you access to the full breadth of both programs. March 15-18. Details and Registration Info >>

More >>

Job Board

New on the Job Board this week:

🦠 Coronavirus Data Specialist (remote, short-term)
Basketball Analytics Intern with the Los Angeles Lakers 🏀
Summer Internship at RStudio (remote)
Data Developer at Songspace
ML Engineer at Stripe (remote)

Check these out and more >>

Data Elixir is curated and maintained by @lonriesberg. For additional finds from around the web, follow Data Elixir on LinkedIn, Twitter or Facebook.