The Prediction and Feature Importance Investigation in Titanic Survival Prediction

Chutong Huang

2024

Abstract

Predicting the survival of Titanic passengers is one of the topics scientists are focusing on. This paper explores the use of the Random Forest (RF) algorithm on a Titanic dataset and analyses the key features that influence the predictions. The RF algorithm is applied to a processed dataset. Feature importance scores are returned for each feature to demonstrate how much it is related to the survival prediction, and the scores are then analyzed in their historical context. Age, fare and sex were found to be the three most significant features in predicting survival. Age is significant as its correlation to survivability, the children are determined to live while the elderly are unable to survive. Fare is a crucial attribute since it is correlated with passenger class, meaning that those paying more are given better information and location to survive. Sex is important because women and children are given priority to survival, while men don’t have that chance. The application of Random Forests shows how well Artificial Intelligence (AI) algorithms can predict problems and spot significant patterns in complex data sets. And the analysis could have useful implications for improving predictive models in other areas where attributes are crucial.

Paper Citation

in Harvard Style

Huang C. (2024). The Prediction and Feature Importance Investigation in Titanic Survival Prediction. In Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML; ISBN 978-989-758-754-2, SciTePress, pages 60-63. DOI: 10.5220/0013487400004619

in Bibtex Style

@conference{daml24,
author={Chutong Huang},
title={The Prediction and Feature Importance Investigation in Titanic Survival Prediction},
booktitle={Proceedings of the 2nd International Conference on Data Analysis and Machine Learning - Volume 1: DAML},
year={2024},
pages={60-63},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013487400004619},
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 - The Prediction and Feature Importance Investigation in Titanic Survival Prediction
SN - 978-989-758-754-2
AU - Huang C.
PY - 2024
SP - 60
EP - 63
DO - 10.5220/0013487400004619
PB - SciTePress