Salifort motors (a fictional company) has high employee turnover recently. With the given data, we are about to explore the reasons that might have caused the employee turnover and suggest ways to improve employee satisfaction and retain them long term.
This repository contains a jupyter notebook with the steps taken to solve the problem, the PACE (Plan, Analyze, Construct, Execute) strategy document and the executive summary of the analysis and model results.
We have given a set of features related to the existing employees with the target variable indicating if the employee has left. Dataset is not added to the repository, since this is a paid course)
Explored the given set of variables and found few of the variables that can be directly used in the model.
Built a Logistic Regression and an XGBoost with the existing features and also engineered some new features and built the model. All the steps taken are given in the Jupyter notebook.
XGBoost model with new features added got 94.5% of F1 score and produced some important features at the top from which insights were derived. Suggestions are given to the HR department. Executive Summary is at Executive_summary