— Insight —
There has been a huge demand for data scientists, analysts, and engineers in the past decade. Is that about to change?
Because of a statistical quirk called “collider bias,” the criminal justice system may be even more racially biased than studies suggest. Here's how collider bias works, including charts that clearly show the problem.
— Tools and Techniques —
STUMPY is a powerful and scalable Python library that efficiently computes something called the matrix profile, which can be used for a variety of time series analysis tasks. This series of posts explores how it works and how you can leverage it for modern time series data mining.
Nice thought experiment that shows how machine learning can amplify bias in a dataset and the hazards that could result.
If you use Google Colab, this is a great collection of lesser-known features that will help boost your productivity.
Lessons learned building a profitable algorithmic trading system using Reinforcement Learning techniques.
— Resources —
This exhaustive Python cheatsheet has been making its way around the web and has recently been updated with sections for Pandas and Plotly.
— Data Viz —
Technical guide for developing data visualizations in Observable Notebooks and then user-testing them on Mechanical Turk.
In this companion article to her Getting Started with Data Viz Journal, Diana MacDonald shares her learning process, including an extensive collection of curated resources. This is well organized and thorough.