This was an Analytics Vidhya Competition.
The goal is to employ a data-driven approach to predict if a particular employee should be promoted or not. Predictions will be based on a couple of features related to the employees.
The problem of competition is formulated as a binary classification problem. The evaluation metric for the competition is F1 score.
Steps Involved: Understanding and Preprocessing the Data, Model Building and Evaluation, SHAPLEY values for interpretability and Ensemble Learning.
Soft Voting has given us the best F1 score (~0.47) of all models.
Please go through Documentation.pdf in order to understand file contents of this repository.
Code is in 656_HR_Analytics.ipynb, and the Competition report is in HR Analytics.pdf.
Thank you for reading, and have a good day!