HR Analytics: Using Machine Learning to Predict Employee Turnover

Matt Dancho, Founder of Business Science, presents how to use machine learning to predict employee turnover at the EARL Boston 2017 conference. He discusses two cutting edge techniques: H2O and LIME. The new H2O automated machine learning algorithm is used to get high accuracy models. The LIME package is used to explain complex deep learning models.

Code accompanying the presentation can be found in the original Business Science blog article: http://www.business-science.io/busine...

Presentation materials & slides can be found at the Business Science GitHub site / presentations repo here: https://github.com/business-science/p...

Packages used: h2o, lime

[LAUNCHING in 2020] Advanced Time Series Forecasting in R course. Become the times series domain expert in your organization. Get notified about this course when it comes out here: https://mailchi.mp/business-science/t...

[Weekly R-tips] Every Tuesday receive a new, worked out R example using practical datasets. Whether it's shortcodes, how to use newly released libraries and functions, and other ways to help you be more efficient and "in the know" Sign up here: https://learn.business-science.io/fre...

[Learning Labs] Every two weeks I host a FREE & LIVE webinar on advanced, intermediate, or cutting-edge topics in data science, Shiny app development, or machine learning. Always includes a code demo. Get notified so you don’t miss out: https://mailchi.mp/business-science/w...