Google Advanced Data Analytics Capstone Project by Markku Laine (2023)
Salifort Motors (a fictional company) is currently experiencing a high turnover of employees, suggesting that a large proportion of its employees may be dissatisfied with their jobs. The goal of this project is to help the company to increase employee retention by conducting an in-depth analysis of the employee survey data and building different machine learning models—Decision Tree, Random Forest, XGBoost—to predict employee attrition.
This project uses an HR dataset from Kaggle. The dataset contains 14,999 rows and 10 columns, each row representing self-reported information from employees.
- The analysis identified high workloads, limited career advancement opportunities, and burnout among high performers as the three main factors behind employees leaving the company.
- In the model comparison,
XGBoost v2outperformed others, achieving an excellent recall rate of 93.5% on the test data in identifying leaving employees. - workload (engineered), tenure, evaluation_score, and satisfaction_level were the four most important features for the model. Surprisingly, promoted and overtime (engineered) were consistently among the least important features for all models.
The performance of the XGBoost v2 model could be potentially improved by (1) collecting more data, such as office location in future surveys and (2) optimizing the model’s decision threshold to attain the highest possible recall score.
Salifort Motors faces a significant employee turnover challenge, with an attrition rate of 16.6%. The analysis identified three main factors contributing to this trend: high workloads, limited career advancement opportunities, and burnout among high performers.
To address these concerns, we recommend a comprehensive strategy that includes:
- Workload balancing: Distribute projects equitably to reduce excessive workloads.
- Promotion strategies: Implement structured promotion plans to identify and support high performers.
- Well-being initiatives: Foster a culture of employee well-being and appreciation through flexible work arrangements, stress management programs, and meaningful recognition.
- Proactive attrition mitigation: Deploy the
XGBoost v2model, which achieved an outstanding recall score of 0.9347, to identify employees at risk of leaving and implement preventative measures.
By implementing these recommendations, Salifort Motors can effectively combat employee turnover, enhance employee satisfaction, and cultivate a thriving workforce.
- Project Proposal [PDF] [DOC]
- Executive Summary [PDF] [PPT]
- Jupyter Notebook [PDF] [JNB]
- HR Dataset [CSV]
- Models [directory]
Install Miniconda, after which we can create a dedicated Python virtual environment for the project:
# Create a virtual environment for Python 3.10
conda create -n gada-capstone-project python=3.10
# Activate the environment
conda activate gada-capstone-project
# Deactivate the environment, if needed
conda deactivateInstall Python dependencies, including those needed for development:
pip install -r requirements-dev.txtNote: Make sure the Python virtual environment is active before installing requirements.
Open the Jupyter notebook:
jupyter notebook employee_retention_project.ipynbThis project comes with pre-commit hooks that include several utility tools to ensure a consistent and high-quality coding style. To install and use (or uninstall) pre-commit, run:
# Install pre-commit into your git hooks
pre-commit install --install-hooks --overwrite
# Run all pre-commit hooks against all tracked files
pre-commit run -a
# Run all pre-commit hooks against all files
pre-commit run --files **/*.py
# Uninstall pre-commit from your git hooks, if needed
pre-commit uninstallCopyright (c) 2023 Markku Laine
This software is distributed under the terms of the MIT License. See LICENSE for details.