CK Logo
Post main image

Should I stay or Should I Go...

Predicting Employee Turnover

One of the latest projects I endeavor on was looking at a Human Resource dataset to make inferences on what factors contribute most towards employee turnover. Running a variety of models, I found my Random Forest Classifier model returned the best F1-score of 87%, explaining how low "Satisfaction Levels" and large counts of "Number of Projects" were the top features to why an employee may have left the company.