Titanic Survival Prediction

This repository contains a complete ML project example: model training, REST API, web interface, and deployment via Docker Compose.

Project Structure

├── code
│   ├── datasets         # source datasets
│   ├── models           # model training scripts
│   └── deployment
│       ├── api          # FastAPI server for the model
│       └── app          # Streamlit web application
├── data                 # (optional) for storing data
├── models               # serialized trained model

How to Run

1. Clone the repository

   git clone <repo_url>
   cd PMLDL_Assignment_1/code/deployment
   

2. Start services via Docker Compose

   docker-compose up --build
   

   This will start two containers: API (FastAPI) and web app (Streamlit).

3. Open the web interface
   In your browser, go to:

   http://localhost:8501
   

   Enter passenger parameters and get the prediction.

---

Note: If you want to retrain the model, run python train_model.py in code/models/ and replace the file in models/.

Parameter Description

• Pclass — ticket class (1, 2, 3)
• Sex — gender (male/female)
• Age — age
• SibSp — number of siblings and spouses aboard
• Parch — number of parents and children aboard
• Fare — ticket fare
• Embarked — port of embarkation (S — Southampton, C — Cherbourg, Q — Queenstown)

Dependencies

All dependencies are listed in the requirements.txt files inside the api and app folders.

Contacts

For questions and suggestions, create an issue or contact via Telegram - @clotjh