This project is a web-based stock price predictor built using LSTM (Long Short-Term Memory) neural networks and Streamlit. It takes historical stock data and forecasts future trends using deep learning. The app allows users to input a stock symbol and visualize price trends alongside moving averages and predictions.
- Fetches live historical stock data from Yahoo Finance
- Preprocesses and scales the data using MinMaxScaler
- Visualizes:
- 50-day, 100-day, and 200-day moving averages
- Actual vs Predicted stock prices
- Uses a stacked LSTM model with dropout layers for robust prediction
- Built with a clean and interactive Streamlit interface
- Frontend: Streamlit
- Backend: Keras (TensorFlow backend), Numpy, Pandas
- Visualization: Matplotlib
- Data Source: yFinance API
- Model: LSTM Neural Network