A Comparison of Machine Learning Techniques for Prediction of Survival using Titanic Dataset

Abstract

This study leverages machine learning algorithms to predict the survival of passengers based on the "Titanic: Machine Learning from Disaster" dataset provided by Kaggle.The dataset is split into training and test sets.The dataset includes various attributes such as Passenger Class, Age, Gender, and Fare.By applying data preprocessing techniques like imputation, label encoding, and feature scaling, a robust model is developed using the training dataset.This paper compares the performance of the Decision Tree algorithm with other advanced models like XGBoost and CatBoost, evaluating their accuracy in predicting passenger survival.The study highlights the importance of feature selection and preprocessing in enhancing model performance, with CatBoost achieving the highest accuracy of 81.82%.The findings contribute to understanding how different factors influenced survival rates, offering insights into the efficacy of various machine learning approaches in predictive analysis.