add readme

README.md

@@ -0,0 +1,50 @@
+---
+title: SKS Course Enrollment Predictor
+emoji: 📊
+colorFrom: blue
+colorTo: green
+sdk: gradio
+sdk_version: 4.44.1
+app_file: app.py
+pinned: false
+---
+
+# SKS Course Enrollment Prediction System
+
+Predicts student enrollment for elective courses using Prophet time series forecasting with semester-specific analysis.
+
+## 🎯 Features
+
+- **Semester-Specific Predictions**: Separate predictions for Semester 1 (Ganjil/Odd) and Semester 2 (Genap/Even)
+- **Time Series Forecasting**: Uses Facebook Prophet for accurate enrollment predictions
+- **Historical Backtesting**: Validates model accuracy with MAE and RMSE metrics
+- **Automated Recommendations**: Suggests which courses to open based on predicted demand
+- **Private Data**: Loads enrollment data from private Hugging Face dataset
+
+## 🚀 How It Works
+
+1. **Select Target Year and Semester**: Choose the academic period to predict
+2. **Generate Predictions**: AI analyzes historical enrollment patterns
+3. **View Recommendations**: See which courses should be opened and recommended quotas
+4. **Review Metrics**: Check model performance (MAE, RMSE)
+
+## 📈 Prediction Strategy
+
+The system uses multiple forecasting strategies:
+- **Prophet Logistic Growth**: For courses with sufficient historical data
+- **Trend-Based Fallback**: For courses with unrealistic Prophet predictions
+- **Mean Fallback**: For courses with limited history
+- **Cold Start**: For new courses without historical data
+
+## 🛠️ Technical Stack
+
+- **Framework**: Gradio for UI
+- **ML Model**: Facebook Prophet
+- **Data Processing**: Pandas, NumPy
+- **Deployment**: Hugging Face Spaces
+
+## 📊 Model Performance
+
+The model is validated through backtesting on historical data:
+- Mean Absolute Error (MAE): ~31 students
+- Root Mean Squared Error (RMSE): ~49 students