🔐 Password Strength Detection using Machine Learning

🌐 Live Demo

• 🔗 Frontend (Streamlit App): https://password-strength-ml.streamlit.app/
• 🔗 Backend API (FastAPI on Render): https://password-strength-ml-m21m.onrender.com
• 📘 API Docs: https://password-strength-ml-m21m.onrender.com/docs

---

📌 Overview

This project is a full-stack Machine Learning application that evaluates password strength in real time.

It combines:

• 🧠 A trained ML model (Random Forest)
• ⚡ A REST API built with FastAPI
• 🎨 An interactive UI built with Streamlit

Users can input a password and instantly get:

• Strength classification (Weak / Medium / Strong)
• Security insights (entropy, patterns)
• Improvement suggestions

---

🚀 Features

🧠 Machine Learning

• Feature-engineered password analysis
• Entropy-based complexity measurement
• Random Forest classifier (multi-class)

⚡ Backend (API)

• REST API using FastAPI
• JSON-based prediction endpoint (/predict)
• Deployed on Render

🎨 Frontend (UI)

• Interactive UI using Streamlit
• Password generator 🔁
• Strength score (0–100)
• Entropy calculation 📊
• Real-time feedback & suggestions
• API health monitoring

---

📂 Project Structure

password-strength-ml/
│
├── assets/ # Screenshots
├── data/ # Dataset (rockyou.txt)
├── models/ # Trained model (.pkl)
│
├── src/
│ ├── data_loader.py
│ ├── preprocessing.py
│ ├── feature_engineering.py
│ ├── model.py
│ ├── train.py
│
├── api/
│ └── main.py # FastAPI backend
│
├── streamlit_app.py # Streamlit frontend
├── main.py # CLI prediction
├── requirements.txt
└── README.md

---

📊 Dataset

Download RockYou dataset from Kaggle and place it in:

data/rockyou.txt

• RockYou password dataset (real-world leaked passwords)
• Synthetic strong passwords generated programmatically

⚠️ Note: Dataset contains weak passwords, so strong passwords are artificially generated to balance the dataset.

---

🧠 Feature Engineering

Password length Uppercase / lowercase count Digits & special characters Character ratios Entropy (randomness measure) Weak pattern detection ("123", "password") Repetition patterns

---

📊 Model Details

Algorithm: Random Forest Classifier Library: scikit-learn Classes: 0 → Weak 1 → Medium 2 → Strong

📊 Model Performance

📈 Performance Accuracy: ~91% F1 Score: ~0.89

---

⚙️ Installation

1. Clone the repository

git clone https://github.com/prathameshsail72-hue/password-strength-ml.git cd password-strength-ml

python -m venv venv venv\Scripts\activate # Windows

pip install -r requirements.txt

Run Backend (API) uvicorn api.main:app --reload

Visit: http://127.0.0.1:8000/docs

Run Frontend (Streamlit) streamlit run streamlit_app.py

---

📈 Evaluation

The model is evaluated using:

• Accuracy
• Precision / Recall / F1-score

---

⚠️ Limitations

No breach detection (e.g. leaked passwords) Rule-based + ML hybrid (not attack simulation) Synthetic strong passwords used for balancing Render free tier causes cold start delays

---

📸 Demo

Password Strength Predictions

Training Output

Project Structure

🌐 Web App Demo (Streamlit)

Run the app locally:

bash streamlit run app.py

🔮 Future Improvements

🔍 Integrate Have I Been Pwned API 🧠 Deep Learning (LSTM / Transformers) 📊 Confidence score output ⚡ Caching & performance optimization 🌍 Multi-language password analysis

---

🤝 Contributing

Contributions are welcome! Feel free to open issues or submit pull requests.

---

📜 License

This project is open-source and available under the MIT License.

---

👨‍💻 Author

Prathamesh Sail GitHub: https://github.com/prathameshsail72-hue