Stock Prediction is built with these core frameworks and modules:
- Streamlit - To create the web app UI and interactivity
- LSTM - To build the Long Short Term Memory model
- Plotly - To create interactive financial charts
- Keras - To Train the Model layer by layer.
- Tensorflow - To Primarily Build & Train LSTM Neural Network model.
The app workflow is:
- User feeds the CSV file.
- Historical data is fetched with CSV file.
- LSTM model is trained on the data
- Model makes multi-day price forecasts
- Results are plotted with Plotly
- Financial charts - Interactive historical and forecast charts
- LSTM forecasting - Make statistically robust predictions
- Backtesting - Evaluate model performance
- Responsive design - Works on all devices
- Clone the repo
git clone - Install requirements
pip install -r requirements.txt- Change directory
cd app.py- Run the app
streamlit run app.pySome potential features for future releases:
- More advanced forecasting models like Transformer
- Quantitative trading strategies
- Portfolio optimization and tracking
- Additional fundamental data
- User account system
- Real Time Data Fetching
"""
This is not financial advice! Use forecast data to inform your own investment research. No guarantee of trading performance. """