Surviving the Titanic tragedy: A sociological study using machine learning models

Introduction

• The Titanic was a symbol of British supremacy and was considered unsinkable.
• It sank on April 15, 1912, after hitting an iceberg, resulting in the deaths of two-thirds of its passengers and crew.

Data and Methodology

• Data included 890 passenger samples for training and 418 for testing.
• Machine learning algorithms, specifically logistic regression and random forest, were used to analyze the data.
• Key variables included title, sex, fare, class, and family density.

Hypotheses

1. Economic Determinants: First-class passengers had a higher chance of survival due to better access to lifeboats and information.
2. Natural Determinants: Younger people had a higher chance of survival.
3. Social Determinants: Women and children had a higher chance of survival due to social norms.

Results

• Title and Sex: These were the most significant factors in determining survival.
• Class and Fare: First-class passengers had a higher survival rate.
• Family Size: Smaller families had a higher chance of survival.
• Age: Younger passengers had a better chance of survival.

Statistical Analysis

• Logistic regression and random forest models were used to predict survival.
• The models showed that title, sex, and fare were the top three variables influencing survival.

Conclusions

• The lack of lifeboats was a major reason for the high death toll.
• Social norms played a significant role in survival, with women and children being prioritized.
• Economic status and proximity to lifeboats also influenced survival rates.