— In the News —
COVID-19 is especially lethal for older adults and those with other health problems. Young adults are more likely to recover but tend to be big spreaders of the disease. That creates two conflicting priorities for the mathematicians. Should they prevent deaths or slow transmission?
— Tools and Techniques —
Data is rarely static and for the most part, your tests shouldn't be either. This is a super useful post that explores the issues and introduces a framework for deciding how to test – from static tests that can be written in SQL, to dynamic tests that require statistics or ML.
Great machine learning tutorial using R and live data from the cryptocurrency markets. This is very well done and includes a hosted Jupyter notebook with sections for data prep, visualization, model validation, predictive modeling, evaluation, and more.
Different business models face different challenges and experimentation strategies that work in one model may not work in another. This post on the StitchFix blog starts out with a thought experiment about the foraging strategies of data-savvy squirrels...
Observable recently announced a round of funding which helps ensure its role as a key data visualization platform for the web. As a celebratory response, a lot of Observable Notebooks were shared around the web this week. This one, in particular, captures why Observable is worth knowing. See the home page for a gallery of examples.
Alibi Detect is an open source Python library focused on outlier, adversarial and drift detection. The package aims to cover both online and offline detectors for tabular data, text, images and time series.
— Resources —
Barça Innovation Hub's annual conference showcases the present and future of statistical modeling and analysis of sports data, especially as it pertains to ⚽. Scroll down the page and check out the research paper finalists for insights into the opportunities of this rapidly evolving field.
— Data Viz —
If you're planning a big Thanksgiving this week, this dashboard shows the likelihood that you'll be sharing dinner with someone who has COVID-19. In the scheme of things, dinner might not be your biggest concern but adding air travel looks like a recipe for disaster.
— Career —
If you find yourself thinking that part of this post doesn't apply to you, "then chances are that is where you are weak at data science."