Absenteeism Prediction System (Apparel Industry)

A machine learning-powered system designed to forecast absenteeism trends and classify employee risk in the apparel industry. This tool empowers HR and management with actionable insights for proactive workforce planning.

🚀 Features

• Employee Risk Classification: Categorizes employees into High, Moderate, and Low risk based on historical absence patterns and demographics using a Random Forest Classifier.
• Smart Forecasting: Predicts daily absence counts for the next 7-30 days using advanced times-series models (LSTM, XGBoost, Gradient Boosting, and Hybrid models).
• Interactive Dashboard: A user-friendly web interface built with Streamlit to visualize risk distribution and forecasting trends.
• Power BI Support: Includes a guide to visualize the processed data in Microsoft Power BI.

🛠️ Tech Stack

• Python 3.13
• Machine Learning: scikit-learn, xgboost, tensorflow (Keras)
• Data Processing: pandas, numpy, openpyxl
• Visualization: matplotlib, seaborn, streamlit

📂 Project Structure

├── processed/              # Processed CSV data files
├── models/                 # Saved ML models (.pkl, .keras)
├── scripts/
│   ├── process_data.py     # Data cleaning and feature engineering
│   ├── train_models.py     # Basic model training (RF)
│   ├── train_advanced_models.py # Advanced training (LSTM, XGBoost)
├── docs/
│   └── powerbi_setup.md    # Guide for Power BI Dashboard
├── app.py                  # Streamlit Dashboard application
├── requirements.txt        # Python dependencies
└── run_app.bat             # Shortcut to run the dashboard

⚙️ Installation

1. Clone the Repository:

   git clone https://github.com/ashenjaygit/Absenteeism-Prediction-System.git
   cd Absenteeism-Prediction-System

2. Install Dependencies:

   pip install -r requirements.txt
   pip install xgboost tensorflow

▶️ Usage

1. Run the Dashboard

You can easily start the application by double-clicking run_app.bat or running:

streamlit run app.py

2. Retrain Models (Optional)

If you have new data in Absenteeism 1.xlsx or daily_absence_summary.xlsx:

# 1. Process new data
python scripts/process_data.py

# 2. Train models
python scripts/train_advanced_models.py

📊 Model Performance

The system evaluates multiple forecasting models and automatically selects the best one. Current performance metrics (MAE):

• LSTM (Deep Learning): ~2.60 (Best)
• Hybrid (XGB+LSTM): ~2.78
• Gradient Boosting: ~2.82
• XGBoost: ~3.13