ISSUE 382  ·   April 12, 2022

Trends

AI could help college baseball players reach the majors, but...

Between wearables, apps, and captured video, there's a lot of data being collected on baseball players. Even video captured on a mobile phone can be used to analyze 2,000-plus types of data including velocities, accelerations and joint positions! The capabilities are impressive but in a world that's increasingly watching us all, the tech is rife with concerns.
Protocol | Kate Kaye

The Data Team’s Guide for Marketing Metrics

Marketing leaders rely on data to meet their quarterly targets. Data teams can use this starter guide to better understand which metrics can accelerate marketing goals and increase ROI. Download Mode’s guide and put it into play with a 14-day Mode trial. 
// sponsored

Tutorials, Projects & Opinions

Real World Recommendation System

Nice introduction to how recommendation systems work at scale. For more, there's a good discussion about this post on Hacker News >>
Fennel AI | Nikhil Garg

Uncommon advice on becoming a data scientist in the public interest

Good advice here for those aspiring to work with data in the public sector. Especially when you're starting out, understanding a domain is more important than developing some idolized set of technical skills.
Hertie School | Alex Engler

Proxycurl API - Scrape LinkedIn profiles in real-time

Thousands of data scientists use Proxycurl API to fetch fresh LinkedIn Profiles to keep their data model up to date. New data teams purchase LinkDB, our database of 600M LinkedIn profiles to bootstrap their AI models. Try Proxycurl API for free today.
// sponsored

Resources

Reproducible and Trustworthy Workflows for Data Science

These notes for a new course at the University of British Columbia are comprehensive and well-organized. Covers a wide range of topics in reproducible and trustworthy workflows including things like version control, virtual environments, managing dependencies, code testing, creating reproducible reports, continuous integration, and much more.
University of British Columbia | Tiffany Timbers

🐙 ML Notebooks

Great series of notebooks that introduces a variety of machine learning tasks and applications. The notebooks are minimal and intended to be easily reusable and extendable. This is an evolving project and there's more to come.
GitHub | Elvis Saravia and and Ritvik Rastogi

Code & Tools

Quarto

Quarto is a new open-source publishing system that lets you create dynamic scientific and technical content using Python, R, Julia, and Observable. You can also author documents as plain text markdown or as Jupyter notebooks and then publish in HTML, PDF, Word, ePub, and lots more. There's great documentation and a gallery of examples here.
quarto

PyGOD

PyGOD is a Python library for graph-based outlier detection algorithms. The package is easy to use and includes a variety of current algorithms. 
GitHub | PyGOD Team

Outlier

DALL·E 2

DALL·E 2 is a new AI system that can create realistic images and art from a description in natural language. The results are super impressive and images have been bouncing around the Internet like wildfire. This page on OpenAI's site includes lots of examples, a short intro video, a link to the research paper and much more.
Open AI

How did you like this issue of Data Elixir?

👎    1     2     3     4     5   👍

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