Logistic Regression Model Used to Predict Employee Attrition
This project contains a logistic regression model with variables pulled from the IBM Database found on Kaggle. For additional information on this dataset, see link included in the Data Used folder.
The model was tested and trained using different data/ The results show that the model is approximately 73.6% accurate (with a threshold of 60%). Within testing, 47 employees had were labeled as individuals that had left the compant. Of those employees, the model was successful in predicting that 40 of them would leave (85.1% accurate). For the complete confusion matrix, see the Building the Model.ipynb file attached in this repository.
https://www.kaggle.com/pavansubhasht/ibm-hr-analytics-attrition-dataset