This is a group project in the Data Science for Business I course where we took a data-driven approach to foster employee retention and enhance operational efficiency by building predictive models on Python.
Employee churn poses multiple organisational challenges. Frequent employee turnover disrupts organisational stability and incurs substantal costs in hiring and training. Therefore, the aim is to identify and mitigate factors that lead to increased employee departure.
- Cost management: reducing churn decreases expenses related to recruitment, onboarding, and training.
- Stability & productivity: retaining employees enhances operational stability and preserves institutional knowledge.
- Proactive retention: understanding churn drivers enables proactive measures to enhance employee satisfaction and retain talent.
Construct a model to predict employee departure likelihood utilising various predictors.
Employee satisfaction, salary, and average monthly hours are anticipated to be significant indicators of turnover.
- Descriptive: utilise data visualisation for initial comprehension of variable distributions and basic relationships.
- Exploratory: execute correlation analysis to discern the factors most influencing turnover.
- Predictive: implement and evaluate classification models at three complexity levels - logistic regression, decision tree, random forests.