Stock Price Prediction System

Description

This is an interactive Stock Price Prediction System built using Streamlit, LSTM (Long Short-Term Memory) neural networks, and Yahoo Finance API. It enables users to:

• Visualize historical stock prices with moving averages.
• Train a custom LSTM model for stock price prediction.
• Evaluate the trained model using various performance metrics.
• Predict future stock prices.

Features

• 📈 Stock Data Visualization: Fetch and visualize stock data from Yahoo Finance.
• 🚀 Train LSTM Model: Train a custom LSTM model on selected stock data.
• 📊 Performance Evaluation: Evaluate the model using RMSE, MAE, and R-squared scores.
• 🔮 Stock Price Prediction: Predict future prices based on trained models.
• 🛠 Download Model & Scaler: Save and reuse trained models for future predictions.

Installation

Prerequisites

Ensure you have Python 3.7+ installed and the required dependencies:

pip install streamlit pandas yfinance plotly numpy keras scikit-learn joblib

Clone the Repository

git clone https://github.com/your-username/stock-price-prediction.git
cd stock-price-prediction

Run the Application

streamlit run app.py

Usage

1. Select a Stock: Choose from available stocks (GOOG, AAPL, AMZN, MSFT) or enter a custom ticker.
2. Set Date Range: Select the start and end dates for analysis.
3. Visualize Data: View stock trends and moving averages.
4. Train the Model: Train an LSTM model on historical stock data.
5. Evaluate Performance: Analyze model accuracy with various metrics.
6. Make Predictions: Use the trained model to forecast future stock prices.

Folder Structure

stock-price-prediction/
│── app.py                # Main Streamlit application
│── train_lstm.py         # LSTM model training script
│── models/               # Saved trained models
│── data/                 # Processed stock data
│── requirements.txt      # Required dependencies
│── README.md             # Project documentation

Contributing

Contributions are welcome! Feel free to fork this repository, make changes, and submit a pull request.

License

This project is licensed under the MIT License.