🎓 Graduate Admission Prediction System

A full-stack machine learning web app to predict your chances of getting into a graduate program, explain feature contributions, and evaluate your Statement of Purpose (SOP) using Gemini AI.

---

🌟 Features

• 🔮 Admission Prediction using a trained ML model
• 🧠 SHAP Explainability for feature impact
• 📝 SOP Evaluation with Gemini AI (via Google Generative AI)
• 📊 Smart Recommendations to improve your profile
• 🏫 University Rating Auto-Fill based on world rank
• 📄 Downloadable summary report

---

🛠 Tech Stack

• Frontend: Streamlit
• Backend: FastAPI
• ML Model: Scikit-learn (Linear Regression)
• Explainability: SHAP
• SOP Evaluation: Gemini (via Google Generative AI)

---

🚀 How to Run Locally

1. Clone the Repository

git clone https://github.com/saishagoel27/GAPS_NTCC
cd GAPS_NTCC

2. Create and Activate a Virtual Environment

python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate

3. Install Requirements

pip install -r requirements.txt

4. Set Up Secrets

Create a secrets.toml file:

GEMINI_API_KEY = "your-google-gemini-api-key"

🔑 Get your API key from: Google AI Studio

5. Run the App

1. source venv/Scripts/activate
2. export GEMINI_API_KEY="your-key-here"
3. chmod +x run_all.sh
4. ./run_all.sh or bash run_all.sh

• FastAPI backend runs on: http://localhost:8000
• Streamlit app runs on: http://localhost:8501

---

📂 Project Structure

graduate-admission/
├── backend/           # FastAPI endpoints
│   ├── main.py
│   ├── utils.py       # Gemini SOP scoring, SHAP, preprocess
│   └── models/        # Trained model .pkl files
├── frontend/
│   └── app.py         # Streamlit UI
├── data/              # Admission dataset, university ranks
├── notebooks/         # Jupyter notebook for training
├── run_all.sh         # Script to run both frontend + backend
├── requirements.txt
└── secrets.toml

---

📸 Project Walkthrough

Project.Walkthrough.mp4

---

📄 License

Under MIT License - free to use and modify

---

🙋‍♂️ Authors

Built by Anishaa and Saisha