Beautiful decision plots in R with ggparty. An entire story in one picture: https://cran.r-project.org/web/packages/ggparty/vignettes/ggparty-graphic-partying.html
Not just for machine learning models: Communication problems are harder than technical problems: https://kdnuggets.us12.list-manage.com/track/click?u=4f2891ebb155b23f120ece0bd&id=28d594ddb0&e=2a6fad5624
https://alastairrushworth.github.io/Exploring-categorical-data-with-inspectdf/
Sometimes, we’re forced to take a look at legacy analytics that make no sense, this gentlemen has a good tutorial on how to do this: https://rud.is/b/2019/06/11/makeover-jumbalaya-beating-dumbbells-into-slopegraphs-whilst-orchestrating-ethercalc/
One of the chief complaints about R is the lack of support for big data projects, but there are workarounds: https://www.r-bloggers.com/three-strategies-for-working-with-big-data-in-r/
What is model calibration? You want to pick parameter values on the basis of external (to your data) evidence, and compare the model’s predictions with the actual data: https://www.r-bloggers.com/an-ad-hoc-method-for-calibrating-uncalibrated-models/
KDnuggets has an interesting article out on building data science teams. Granted, they’re talking about a pretty big team here!: https://www.kdnuggets.com/2019/07/disruptive-data-science-teams-best-practices.html