Human Capital Analytics – Employee Turnover Prediction

This project applies advanced analytics and machine learning techniques to identify the key drivers of employee turnover and develop targeted retention strategies.

📁 Project Structure

Human-Capital-Analytics/
├── code/
│   ├── R-file-3-submittedfile.R
│   └── R-file-3.R
├── CSDA 6010 - Project 1 - Presentation - SubashYadav-1.pptx
├── CSDA 6010 Project 1 Final- Subash Yadav-1.docx
├── CSDA 6010 Project 1 Final- Subash Yadav-1.pdf
├── data/
│   └── Employee.csv
├── Human Capital Analytics For Students.docx

🎯 Business Goals

• Identify key drivers of employee turnover such as job dissatisfaction, limited promotion, and high workload.
• Predict turnover risk using data-driven models.
• Segment employees into actionable groups for targeted retention efforts.

📊 Analytical Approaches

• Exploratory Data Analysis (EDA)
• Hypothesis Testing (Salary, Safety, Promotion)
• Feature Engineering:
  • Satisfaction-to-Performance Score
  • Binned Workload Categories
• Predictive Modeling:
  • Logistic Regression (AUC = 0.84)
  • Decision Tree (Accuracy = 96.33%)
  • Random Forest (Accuracy = 97.71%)
• Clustering using K-means (3 Segments)

📈 Key Insights

• Satisfaction Level is the most critical predictor of turnover.
• Employees with low satisfaction and high tenure are most likely to leave.
• Workload imbalance and lack of promotions also drive attrition.
• Clustering revealed:
  • Cluster 1: High satisfaction, low risk
  • Cluster 2: Moderate satisfaction, medium risk
  • Cluster 3: Low satisfaction, high tenure, high risk

🛠 Tech Stack

• R (caret, glmnet, cluster)
• CSV Dataset: Employee.csv
• Visualization Tools: ggplot2, base R plots

📎 Deliverables

• Full analytical report with visuals (PDF)
• Cleaned code scripts
• Derived features for improved interpretability
• Final model and clustering insights

📬 Author

Subash Yadav
LinkedIn
GitHub