A Streamlit-based web application for stock price prediction using Machine Learning, with interactive data visualization and a portfolio trading simulator.
Project context: I built SMINT during my 2nd trimester of my BSc in Data Science as part of my Object-Oriented Programming (OOP) course.
This was one of my first complete end-to-end projects, focused on practicing clean structure, core OOP concepts, and building a working ML-powered app.
I’m continuously improving and learning — future versions will include stronger modeling, validation, and more advanced features.
- SQLite-based user registration and login
- SHA-256 password hashing for security
- User profile management
- Upload stock data (CSV, Excel, TXT)
- Multiple chart types: Line, Bar, Scatter, Area, Candlestick
- 7-day and 30-day Moving Averages
- Volume chart overlay
- Interactive date range filtering
- Data statistics (Min, Max, Mean, Median, Std Dev)
- Download filtered data as CSV
- ML-based stock price prediction using trained model
- Single date prediction
- Date range prediction with visualization
- Predicted vs Actual price comparison chart
- Prediction accuracy metrics
- Prediction history tracking
- Virtual portfolio with $100,000 starting cash
- Buy/Sell stocks
- Real-time portfolio tracking
- Transaction history
- Performance summary
- Clone the repository
git clone https://github.com/Az-main/SMINT-Stock-Prediction.git cd SMINT-Stock-Prediction