ISSUE 384  ·   April 26, 2022

Papers

Null Island, the most real of fictional places

Null Island is a bustling destination spot that you've probably never heard of. It has vacation rentals, restaurants, an airline, and social media activity with millions of uploaded photos. But in the real world, Null Island doesn't exist. It's a digital artifact that's been largely seen as a joke amongst GIS researchers. This paper explores its existence, its evolution, and the implications for the data community.
Levente Juhasz, Peter Mooney

Sponsored Link

Understanding and Overcoming Four Types of Biases in AI

There are four types of biases found in machine learning models. These are algorithmic bias, sample bias, prejudicial bias, and measurement bias. How do each of these biases arise and how are each of them mitigated? Read this article to understand how you can produce better business outcomes by training AI models to do precisely what they are meant to do.

Tutorials, Projects & Opinions

Creating Confidence Intervals for ML Classifiers

Including confidence intervals to address uncertainty is standard practice, but it is not straightforward how to pick the best approach. This is a comprehensive post that explores several ways to calculate confidence intervals and when to use each one.
Sebastian Raschka

Porting a distill blog to quarto

Quarto is a new open-source publishing system that lets you publish dynamic documents in a wide variety of formats including HTML, PDF, Word, ePub, and more. In this post, Danielle Navarro walks through the process of moving a blog to Quarto and along the way, she explores Quarto's features and quirks, how to work with it, and where it shines.
Danielle Navarro 

Data tests and the broken windows theory

Data tests are here to stay, but they can be challenging to get right. We've only scratched the surface of their full potential, and this article explores common issues of data tests and how to overcome them.
Mikkel Dengsøe

How to build more meaningful machine learning models with third-party data

Join this webinar to discover how to build more meaningful machine learning models with third-party data from AWS Data Exchange. You’ll learn how to improve forecasting confidence and deliver effective insights with guidance from industry leaders. Register Now
// sponsored

Code & Tools

performance

A crucial aspect when building a regression model is evaluating the quality of model fit. The open-sourced performance package fills this gap by providing utilities for computing indices of model quality and goodness of fit. It uses measures such as r-squared, RMSE, ICC, and checks for singularity, convergence, and more.
GitHub | easystats

Cog: Containers for machine learning

Cog is an open-source tool that lets you easily package machine learning models in a standard, production-ready container. With Cog, you get docker containers with defined inputs and outputs, an automatic RESTful HTTP API, a queue worker, and more.
GitHub | Replicate

Career

Take Control of Your Desk

This comprehensive guide to career growth is tailored towards ICs, managers, and founders who are working in fast moving startups. Includes tips for individual career planning, talent retention, and ideas for developing career ladders.
First Round Review

How did you like this issue of Data Elixir?

👎    1     2     3     4     5   👍

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