ISSUE 400  ·   August 16, 2022

Discussions

Venture Capital for Data Science 

Have a data idea you'd like to turn into a business? This is a great discussion with Sarah Catanzaro about the investing landscape and what it takes to go from idea to funding. Sarah is a General Partner at Amplify Partners, where she focuses on data-oriented startups. This is a good listen and there's also a complete transcript if you prefer to read.
Super Data Science | Jon Krohn with Sarah Catanzaro

Sponsored Link

Twitter Spaces: Responsible and Trustworthy AI

Listen in on a live Twitter Spaces at 9:30am PST on Wednesday, August 17, featuring Ben Lorica of Gradient Flow, Andrew Burt of bnh.ai, and Bob Friday of Juniper Networks. The conversation will discuss what it takes to get to mature adoption of AI in the enterprise, and the role trust plays throughout. 

Tutorials, Projects & Opinions

The Singular Value Decomposition

The singular value decomposition is something of a Swiss Army knife of linear algebra. It's a popular way to compute a principal component analysis and it's also used for linear projections, linear regressions, and much more. This is a fantastic deep dive that shows how it works and the variety of things you can do with it.
Peter Bloem

How the NYT uses ML to make its paywall smarter

When The New York Times first launched its paywall, the meter count was the same for all users. A lot's changed since then and its paywall now uses a causal machine learning model called the Dynamic Meter to set personalized meter limits. Here's how it works.
New York Times Open | Rohit Supekar

Bayesian A/B tests with millions of observations

Data scientists are increasingly using Bayesian methods but at scale, Bayesian methods can be slow. This post walks through the issues and shows how the authors got a proof-of-concept A/B test pipeline to work very fast even with very large numbers of observations.
PyMC Labs | Benjamin Vincent & Maxim Kochurov

Excel is Pretty Dang Cool

Whether you love it or hate it, there's a lot that Excel can do. Here are some lesser known features and if you're in for a good debate, check out the discussion on Hacker News >>
Computer Things | Hillel Wayne

Logistic Regression - Regression for Classification

Logistic regression is a supervised learning algorithm that can be used to classify data into categories, or classes, by predicting the probability that an observation falls into a particular class based on its features. This is a short visual explainer that makes it easy to see how it works.
MLU-Explain | Erin Bugbee & Jared Wilber

Resources

Machine Learning Zoomcamp

Great resource for getting started in Machine Learning. Covers a wide variety of topics with lecture videos, slides, assignments and a Slack channel. The next course begins in September. Free.
DataTalksClub | Alexey Grigorev

Data Visualization

Pyramid Plots in ggplot2

Pyramid plots are often used to compare the ages of different populations. They're intuitive to understand but can be more complicated to build than you might think. This is a nice tutorial that shows a simple way to make them using ggplot2.
Credibly Curious | Nicholas Tierney

Aquarel 🎨

Aquarel is a lightweight templating engine and wrapper for Matplotlibs' rcparams to simplify styling plots. Aquarel templates can be defined programmatically and be serialized and shared in a JSON format.
GitHub | Lukas Gienapp

How did you like this issue of Data Elixir?

👎    1     2     3     4     5   👍

Was this email forwarded to you?  Sign up here >>