This is a Project from the Data Scientist Program Degree by Data Science Academy Brazil. The idea is to build a classification model to predict if an employee will or not leave the company based on the variables satisfaction level, number of projects one is working, number of hours worked in a month and the years this person is in that company.
Learn more about this project in this link to the project report.
Access the Shiny app here: link
Activities performed:
- Data Exploration and Analysis
- Data Visualization (Histograms, Boxplots, Barplots)
- Correlation check
- Feature Engineering with Random Forest to select the best attributes for the model
- Classification: Training and Prediction with Random Forest (99% accurate) and XGBoost (97% accurate)
- (Creation of Random Forest and XGBoost with tidymodels)
- Creation of a Shiny App for user interaction and make predictions.