🏨 Hotel Reservation Cancellation Prediction

App

🚀 Project Overview

This project aims to predict whether a hotel booking will be canceled using a Machine Learning Operations (MLOps) pipeline. It integrates data ingestion, preprocessing, model training, deployment, and a user-friendly web application for real-time predictions. The model is trained on the Hotel Reservations Dataset, which includes various booking attributes influencing cancellations.

🐂 Project Structure

├── artifacts/            # Stored models & processed data
├── config/               # Configuration files (YAML, parameter settings)
├── custom_jenkins/       # CI/CD pipeline setup (Jenkins)
├── notebook/             # Exploratory Data Analysis (EDA) & experimentation
├── pipeline/             # ML pipeline for data preprocessing & training
├── src/                  # Core scripts (ingestion, preprocessing, model training)
├── utils/                # Utility functions for data transformation & logging
├── static/               # Web UI stylesheets
├── templates/            # HTML templates for the web UI
├── logs/                 # Application & model logging
├── venv/                 # Virtual environment setup
├── Dockerfile            # Containerization setup
├── Jenkinsfile           # CI/CD automation pipeline
├── requirements.txt      # Required dependencies
└── setup.py              # Package installation setup

🔄 MLOps Workflow

1. Database & Project Setup: Configure and initialize the required infrastructure.
2. Data Ingestion & Processing: Load, clean, and preprocess hotel booking data.
3. Model Training & Experimentation: Train a LightGBM model with hyperparameter tuning.
4. Versioning & Tracking: Maintain data and code versioning for reproducibility.
5. CI/CD Pipeline: Automate training, validation, and deployment using Jenkins.
6. Web Application: Flask-based UI for real-time cancellation prediction.
7. Deployment: Containerized with Docker and deployed on Google Cloud Run.

🛠️ Getting Started

Prerequisites

Ensure you have Python installed (Recommended: Python 3.8+).

# Clone the repository
git clone https://github.com/SM0311/MLOPS-HOTEL-BOOKING-CANCELLATION.git
cd MLOPS-HOTEL-BOOKING-CANCELLATION

# Create and activate virtual environment
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Run Flask Application
python application.py

🎯 Usage

• Train Model:

  python training_pipeline.py

• Run Web App Locally:

  python application.py

• Docker Deployment:

  docker build -t hotel-predict .
  docker run -p 5000:5000 hotel-predict

⚙️ CI/CD Automation

This project follows a CI/CD workflow using Jenkins:

• Automated data ingestion and preprocessing.
• Continuous model training and validation.
• Containerized deployment via Docker & Google Cloud Run.

📊 Model Performance

The model leverages LightGBM for high accuracy and efficiency. Performance metrics include:

• Accuracy: 90%+
• Precision & Recall: Balanced trade-off for optimal prediction.
• Feature Importance: Identifies key booking attributes influencing cancellations.

🤝 Contributing

Contributions are welcome! Feel free to fork this repository, create pull requests, and report issues.

📩 Contact

For inquiries, reach out via Email: msuraj20@yahoo.com