Stock Ticker Analysis and Prediction System using Stacked LSTM

📋 About

Ticker-Prediction System is an advanced machine learning project that leverages LSTM (Long Short Term Memory) neural networks combined with XGBoost ensemble methods to predict stock market trends using time-series analysis.

Key Features

• 🤖 Deep Learning Architecture: Stacked LSTM model for capturing temporal dependencies in stock data
• 📊 Data Preprocessing: Advanced techniques for continuous dataset normalization and transformation
• 🚀 Ensemble Method: XGBoost integration to enhance prediction accuracy
• 🎨 Dual Interface: Both Streamlit web app and React-based frontend
• 🐳 Docker Support: Containerized deployment with version tracking
• 📈 Time-Series Prediction: Sophisticated forecasting for financial data

Note: All research, testing, and model training work are maintained in a separate repository with different epoch cycles and parameters.

📖 Description

This project combines cutting-edge deep learning techniques with practical deployment strategies:

• Core Technology: LSTM networks with XGBoost ensemble methods for superior prediction performance
• Multiple Interfaces:
  • Streamlit Application - Quick web-based interface
  • React Application - Full-featured frontend with advanced UI/UX
• Containerization: Docker images for consistent deployment across environments with semantic versioning

🚀 Getting Started

Prerequisites

• Python 3.8+
• Node.js 14+
• Docker (optional, for containerized deployment)
• Git

1. Local Installation

Frontend (React)

Navigate to the Ticker Predictor Website folder:

cd "Ticker Predictor Website"
cd "Front-end"
npm install
npm run start

The React app will be available at https://localhost:3000

Backend (Python)

In a separate terminal, navigate to the Backend folder:

cd "Ticker Predictor Website"
cd "Backend"
python waitress_server.py

The backend server will connect with the React frontend automatically.

Visual Guide:

Terminal Output Example:
✓ Backend running on http://localhost:5000
✓ Frontend running on http://localhost:3000
✓ Connected and ready for predictions

2. Docker Deployment

Option A: React + Python Backend

Terminal 1 - Backend:

docker run -p 5000:5000 debasishray/predictor-backend:latest

Terminal 2 - Frontend:

docker run -p 3000:3000 debasishray/predictor-frontend:latest

Then navigate to: https://localhost:3000

Option B: Streamlit Application

docker run -p 8501:8501 debasishray/streamlit-app:latest

Access the app via the link provided in the terminal output.

🐳 Docker Image Deployment to GitHub Packages

Step-by-Step Guide

1. Create a Docker image replica with a new tag:

docker tag debasishray/streamlit-app:v1.0 webapp

2. Tag for GitHub Container Registry:

docker tag webapp ghcr.io/debasishray16/stockpredictor/webapp:latest
docker image ls

3. Authenticate with GitHub Packages:

echo "YOUR_PAT_TOKEN" | docker login ghcr.io -u debasishray16 --password-stdin

4. Push to GitHub Packages:

docker push ghcr.io/debasishray16/stockpredictor/webapp:latest

Format: ghcr.io/<username>/<repository>/<image>:tag

📊 Project Structure

StockPredictor/
│
├── 📄 README.md                          # Project documentation
├── 📄 LICENCE                            # MIT License
├── 📄 .gitignore                         # Git ignore configuration
├── 📄 docker-compose.yml                 # Docker Compose configuration
│
├── 📂 Ticker Predictor Website/          # Main application folder
│   │
│   ├── 📂 Front-end/                     # React Frontend Application
│   │   ├── 📄 package.json               # Node.js dependencies
│   │   ├── 📄 package-lock.json          # Dependency lock file
│   │   ├── 📄 tailwind.config.js         # Tailwind CSS configuration
│   │   ├── 📄 Dockerfile                 # Frontend container config
│   │   ├── 📄 .dockerignore              # Docker ignore rules
│   │   ├── 📄 .gitignore                 # Git ignore rules
│   │   │
│   │   ├── 📂 public/                    # Static public assets
│   │   │   ├── 📄 index.html             # Main HTML entry point
│   │   │   ├── 📄 manifest.json          # PWA manifest
│   │   │   └── 📄 robots.txt             # SEO robots configuration
│   │   │
│   │   └── 📂 src/                       # Source code
│   │       ├── 📄 index.js               # React entry point
│   │       ├── 📄 index.css              # Global CSS styles
│   │       ├── 📄 App.js                 # Main App component
│   │       ├── 📄 App.css                # App-specific styles
│   │       ├── 📄 reportWebVitals.js     # Performance monitoring
│   │       │
│   │       ├── 📂 Components/            # React Components
│   │       │   ├── 📄 Main.js            # Main dashboard component
│   │       │   ├── 📄 Sidebar.js         # Sidebar navigation (70KB)
│   │       │   ├── 📄 Dashboardview.js   # Dashboard view component
│   │       │   ├── 📄 StockPredictionChart.js  # Chart visualization
│   │       │   ├── 📄 LoadingAnim.js     # Loading animation component
│   │       │   ├── 📄 Spinner.js         # Spinner component
│   │       │   ├── 📄 aboutModel.js      # Model information component
│   │       │   ├── 📄 companyInfo.js     # Company details (9KB)
│   │       │   ├── 📄 companyDesc.js     # Company description
│   │       │   │
│   │       │   └── 📂 Styles/            # Component-specific styles
│   │       │       └── 📄 spinner.css    # Spinner animations
│   │       │
│   │       ├── 📂 router/                # Route configuration
│   │       │   └── 📄 Router.js          # React Router setup
│   │       │
│   │       └── 📂 assets/                # Media assets
│   │           ├── 🎞️ idle.gif          # Idle state animation
│   │           ├── 🎞️ loading.gif       # Loading animation
│   │           └── 🎞️ loading2.gif      # Alternative loading animation
│   │
│   └── 📂 Back-end/                      # Python Backend Server
│       ├── 📄 model_api.py               # ML model API (12.7KB)
│       ├── 📄 waitress_server.py         # WSGI server launcher
│       ├── 📄 requirements.txt           # Python dependencies
│       ├── 📄 Dockerfile                 # Backend container config
│       ├── 📄 .dockerignore              # Docker ignore rules
│       │
│       └── 📂 model/                     # Trained ML Models
│           ├── 🤖 8_15_23_125_LXg.h5     # LSTM model (2.2MB)
│           └── 🤖 8_15_23_300_LXg.h5     # LSTM model (2.2MB)
│
├── 📂 assets/                            # Project assets
│   ├── 📂 images/                        # Documentation images
│   │   ├── 🖼️ Website_Preview.png       # Website screenshot
│   │   └── 🖼️ terminal_screenshot.png   # Terminal output screenshot
│   │
│   └── 📂 dataset/                       # Training datasets (optional)

🔗 Related Resources

• Model Training Repository: Stock-Prediction-Models
• Live Streamlit App: Ticker Prediction - Streamlit
• Model Research: Various epoch cycles and parameter configurations available in training repo

👥 Contributors

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🤝 Support & Contributing

For questions, suggestions, or issues:

• Open an Issue
• Check the Model Training Repository for research details

---

Last Updated: 2026-04-29 | Status: Active Development