Multi-layered cybersecurity of databases in CRM systems

Author(s) Collection number Pages Download abstract Download full text
Чорняк В. О., Tymchenko O. V. № 1 (70) 60-71 Image Image

The article explores modeling the probability of survival for Titanic passengers using machine learning and multicriteria decision analysis (MCDA) methods. The aim of the study is to identify key factors that influenced survival outcomes and to develop a predictive model with high classification accuracy.

The analysis is based on an open dataset from the Kaggle platform, which contains detailed information on passenger age, gender, ticket class, fare, family ties, and other relevant features. A complete data preprocessing cycle was carried out, including outlier detection, normalization, encoding of categorical variables, and splitting the dataset into training and testing sets.

The study applied a range of machine learning algorithms: logistic regression, decision trees, support vector machines (SVM), random forest, and gradient boosting. These models were evaluated based on standard metrics to determine their predictive performance. To interpret model outputs and assess feature importance, the LIME and TOPSIS methods were used, allowing for a more transparent evaluation of which variables most influenced survival.

The findings confirm the critical role of socio-demographic factors in survival likelihood, particularly gender, age, and ticket class. Women, children, and first-class passengers had notably higher chances of survival. These insights mirror historical realities and have practical implications for modern risk modeling, especially in high-stakes environments where quick and accurate decision-making is essential.

Integrating MCDA with machine learning enhances both interpretability and reliability of predictive models. The approach can be valuable in designing decision-support systems for transport safety, disaster response, and other areas where forecasting human outcomes is crucial.

Keywords: multicriteria analysis, machine learning, data analysis, forecasting, support vector method, random forest.

doi: 10.32403/1998-6912-2025-1-70-49-59


  • 1. Yu, A. (2023). Predicting Titanic Survivors by Using Machine Learning, Highlights in Science, Engineering and Technology, 34, 360-367, https://doi.org/ 10.54097/hset.v34i.5494.
  • 2. Zhang, Y., & Haghani, A. (2015). A gradient boosting method to improve travel time pre­dic­tion, Transportation Research Part C: Emerging Technologies, 58, 308-324.
  • 3. Titanic disaster prediction based on machine learning algorithms. (n.d.). ResearchGate. https://www.researchgate.net/publication/370569058_Titanic_Disaster_Prediction_Based_on_Machine_Learning_Algorithms.
  • 4. Morokhovych, V., Liakh, I., Khomyak, M., & Morokhovych, B. (2024). Prohnozuvannia pasazhyriv, yaki vyzhyly pid chas katastrofy «Titanika», za dopomohoiu dereva pryiniattia rishen. Computer-integrated technologies: education, science, production, (55), 161-166.
  • 5. Singh, A., Saraswat, S., & Faujdar, N. (2017). Analyzing Titanic disaster using machine learning algorithms. In 2017 International Conference on Computing, Communication and Automation (ICCCA) (pp. 406–411). https://doi.org/10.1109/CCAA.2017.8229835.
  • 6. An overview of multi-criteria decision analysis (MCDA) application in managing water-related disaster events. (n.d.). MDPI Water. https://www.mdpi.com/2073-4441/13/10/1358.
  • 7. Haque, M. A., Shivaprasad, G., & Guruprasad, G. (2021). Passenger data analysis of Titanic using machine learning approach in the context of chances of surviving the disaster. IOP Conference Series: Materials Science and Engineering, 1065(1). https://doi.org/10.1088/1757-899X/1065/1/012042.
  • 8. Singh, K., Nagpal, R., & Sehgal, R. (2020). Exploratory data analysis and machine learning on Titanic disaster dataset. In 2020 10th International Conference on Cloud Computing, Data Science & Engineering (Confluence) (pp. 1–6).
  • 9. Prokhorenkova, L., Gusev, G., Vorobev, A., Dorogush, A. V., & Gulin, A. (2018). CatBoost: Unbiased boosting with categorical features. In Advances in Neural Information Processing Systems (pp. 6638–6648).
  • 10. Analysis of Titanic disaster using machine learning algorithms. (n.d.). ResearchGate. https://www.researchgate.net/publication/353352089_Analysis_of_Titanic_Disaster_using_Machine_Learning_Algorithms.
  • 11. Bogle, S., & Potter, W. (2015). A machine learning predictive model for the Jamaica frontier market. In Proceedings of the 2015 International Conference of Data Mining and Knowledge Engineering.
  • 12. Kaggle. (n.d.). Kaggle. https://www.kaggle.com/.
  • 13. YasserH. (n.d.). Titanic dataset. Kaggle. https://www.kaggle.com/datasets/yasserh/titanic-da­taset.