Titanic Survival Prediction – Streamlit Web App

Overview

This project is a machine learning web application built using Streamlit that predicts passenger survival on the Titanic dataset.
The app allows users to input passenger details and receive a survival prediction based on a trained model.

The project is intended for learning, demonstration, and experimentation with classical machine learning workflows.

Dataset

• Source: Titanic dataset (Kaggle)
• Target variable: Survived
• Features used:
  • Passenger class
  • Sex
  • Age
  • Fare
  • Number of siblings/spouses
  • Number of parents/children
  • Embarked port

Tech Stack

• Python
• Streamlit
• Pandas
• NumPy
• Scikit-learn

Machine Learning

• Data preprocessing (handling missing values, encoding categorical data)
• Model training using classical ML algorithms (e.g., Logistic Regression / Random Forest)
• Model evaluation using accuracy metrics
• Trained model used for real-time prediction via Streamlit UI

Application Features

• Interactive input fields for passenger data
• Real-time survival prediction
• Simple and clean web interface
• Runs locally without external services