Using R, the Tidyverse, H2O, and Shiny to reduce employee attrition
summary
Claims that losing 200 high performers costs about $15M/year (~$78k per high-performer) and shows a deployed Shiny dashboard that predicts individual attrition risk (example employee at 47% risk), built with tidyverse for prep and visualization, H2O AutoML (recipes + H2O) for modeling, and lime for per-employee explanations. The app surfaces actionable levers (stock options, training) for managers, reports reduced attrition from ~20% to ~9% and converts that improvement into millions saved.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 2:10Employee attrition problem
- 9:11Attrition prediction Shiny app demo
- 13:44Shiny apps for business decisions
- 14:07Data science workflow
- 18:47Learning tools and next steps
related talks
01h 39mApplying Rigorous Machine Learning Methods in Business Strategy
20mHow the cloud changed the way I do analytics and research
45mAgility in HR Analytics bringing business benefit
50mScalable Data Science & Deep Learning with H2O
38mRip the Data
27mDeploying distributed AI and machine learning in financial services
14mMLOps at Scale: AI Systems for HR Transformation
18mCross-Industry Anomaly Detection Solutions with R and Shiny