Employee Attrition Analysis

Skills, Tech Stack, and Libraries

1. Skills: Data Wrangling, Predictive Modeling, Statistical Analysis, Data Visualization

2. Tech Stack: Python, SQL, Tableau, Power BI

3. Libraries: Pandas, NumPy, Matplotlib, Seaborn, Scikit-learn

Description and Approach

Objective:

I developed an advanced predictive analytics system to identify the factors contributing to employee attrition and predict employees at risk of leaving the organization. The project aimed to support HR teams in proactive decision-making and improve employee retention strategies.

Approach:

1. Data Collection and Cleaning:

   • Used SQL to extract employee data from HR databases, including demographics, performance metrics, and exit interview data.

   • Preprocessed the data using Pandas to handle missing values, standardize data formats, and encode categorical variables (e.g., job roles, departments).

2. Exploratory Data Analysis (EDA):

   • Conducted EDA using Matplotlib and Seaborn to identify trends in attrition rates across job roles, departments, and demographics.

   • Discovered key patterns such as high turnover rates in specific departments and correlations between performance ratings and attrition.

3. Feature Engineering:

   • Created additional features such as employee tenure, overtime hours, and promotion history to enhance model accuracy.

   • Normalized numerical features and scaled data to improve model performance.

4. Predictive Modeling:

   • Developed a classification model using Scikit-learn (Random Forest and Logistic Regression) to predict employee attrition based on historical data.

   • Fine-tuned hyperparameters to optimize the model’s accuracy, precision, and recall.

Dashboard Design:

Designed an interactive dashboard in Tableau to provide:

1. Attrition heatmaps by department and job role.

2. Key predictors of attrition ranked by importance.

3. Predicted attrition risks for individual employees.

Automation:

Automated the data pipeline to ensure real-time updates, allowing HR teams to act on the latest insights.

Code Flow:

1. Extract and preprocess data using SQL and Pandas.

2. Perform EDA to uncover trends and correlations.

3. Engineer features for predictive modeling.

4. Train and evaluate classification models using Scikit-learn.

5. Visualize results in Tableau and deploy the dashboard for HR teams.

Results

The project delivered a comprehensive attrition analysis system with the following outcomes:

• Improved Retention Strategies: HR teams identified high-risk employees and implemented targeted interventions, reducing attrition by 20%.

• Key Insights: Highlighted critical factors contributing to attrition, such as workload imbalance and lack of career advancement.

• Real-Time Monitoring: The dashboard enabled proactive monitoring of employee sentiment and risk factors.

• Cost Savings: Reduced turnover-related costs by improving retention and streamlining HR interventions.

This project demonstrated the power of predictive analytics in transforming HR operations and enhancing workforce management.

Git Link

For more information and code, visit the Git link.