Titanic Survival Prediction

Project Overview

This project aimed to predict survival outcomes of Titanic passengers using a range of machine learning models. It included data cleaning, feature engineering, and performance comparison of algorithms.

Key Objectives

• Analyze passenger data to uncover survival trends.
• Train machine learning models to predict survival probabilities.
• Evaluate and compare model performance.

Tools & Libraries Used

• R Programming
• Libraries: rpart, e1071, class, dplyr, ggplot2

Results

• Decision Tree: Accuracy - 87%
• Logistic Regression: Accuracy - 80%
• Naive Bayes: Accuracy - 81.8%
• KNN: Accuracy - 85.3%

Project Outcomes

Females and 1st-class passengers had the highest probability of survival. Decision Trees emerged as the best-performing model, with an accuracy of 87%.

Download Report

Download Final Report

Source Code

View on GitHub