Abstract
Predicting passenger survival in the Titanic disaster has become a widely studied classification problem in machine learning. In this study, we apply and compare several machine learning techniques to predict survival outcomes using passenger demographic and socioeconomic characteristics. Model performance is evaluated using prediction accuracy on held-out data. Our results show that the Conditional Forest method achieves the highest predictive accuracy of 81.34%, outperforming other competing models. Consistent with historical accounts, the analysis confirms that women and passengers from higher socioeconomic classes were significantly more likely to survive the disaster. These findings demonstrate the effectiveness of ensemble learning methods for classification tasks and highlight the importance of demographic and socioeconomic factors in survival outcomes. The study illustrates how machine learning techniques can be used to uncover meaningful patterns in historical data and provide interpretable insights into real-world events.
