Survival Analysis of the Titanic Using Random Forests

Tianchong Tang

2024

Abstract

The Titanic disaster is one of the most widely studied maritime tragedies. Analyzing passenger survival rates has become a hot topic. This research endeavors to forecast the likelihood of survival among Titanic voyagers by employing a random forest algorithmic approach. The datasets employed in this analysis include features such as PassengerId, Survived, Pclass, Name, Sex, Age, SibSp, Parch, Ticket, Fare, Cabin, and Embarked. To enhance prediction performance, the author implemented a random forest algorithm, which integrates multiple decision trees. Following data preprocessing, the dataset was randomly separated into a training set, comprising 80%, and a test set, constituting 20%. Across 500 distinct iterations, the data was randomly split into training and test sets. The random forest model achieved an average accuracy of 0.8013, demonstrating its effectiveness in assessing the likelihood of Titanic voyagers' endurance. This underscores the considerable potential of the random forest algorithm in conducting survival analyses.

Paper Citation

in Harvard Style

Tang T. (2024). Survival Analysis of the Titanic Using Random Forests. In Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML; ISBN 978-989-758-754-2, SciTePress, pages 120-123. DOI: 10.5220/0013510300004619

in Bibtex Style

@conference{daml24,
author={Tianchong Tang},
title={Survival Analysis of the Titanic Using Random Forests},
booktitle={Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML},
year={2024},
pages={120-123},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013510300004619},
isbn={978-989-758-754-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML
TI - Survival Analysis of the Titanic Using Random Forests
SN - 978-989-758-754-2
AU - Tang T.
PY - 2024
SP - 120
EP - 123
DO - 10.5220/0013510300004619
PB - SciTePress