Comparative Modeling of Stroke Prediction Using Advanced Machine Learning Techniques

Pereira, Livia Anthony (2024) Comparative Modeling of Stroke Prediction Using Advanced Machine Learning Techniques. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (2MB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

A stroke transpires when there is a sudden cessation of blood flow to a specific region of the brain. This abrupt interruption leads to the gradual demise of brain cells, resulting in disability contingent on the affected brain area. Timely identification of symptoms serves as a crucial factor in providing valuable insights for stroke prediction and facilitating a health-conscious lifestyle. The current research employs machine learning (ML) methodologies to craft and assess multiple models, aiming to construct a resilient framework for the enduring prognostication of stroke occurrences. This research scrabbles into the critical dominion of stroke prediction, employing a comprehensive approach encompassing detailed data preprocessing, innovative feature engineering, and strategic model training. The implementation involves a systematic process, from initial data collection and preprocessing to the application of two distinct feature engineering approaches – correlation analysis and feature importance evaluation. The ultimate model training unfolds with the evaluation of four prominent algorithms: K Nearest Neighbor, Balanced Random Forest, Catboost, and XGBoost. Notably, XGBoost emerges as the most surpassing algorithm, attaining an outstanding accuracy of 98.11%, F1-Score of 0.9811, and Roc Auc score of 0.9977, showcasing its unparalleled efficacy in stroke prediction. The research underscores the importance of diverse feature engineering approaches influencing stroke risk as Feature selection using one-vs-all feature selection surpasses the feature selection by correlation analysis. The findings contribute to the burgeoning field of healthcare analytics and signify a step towards more accurate and targeted stroke prevention strategies.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Chikkankod, Arjun UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science R Medicine > Healthcare Industry Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning H Social Sciences > HM Sociology > Information Science > Communication > Medical Informatics
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	05 Jun 2025 13:05
Last Modified:	05 Jun 2025 13:05
URI:	https://norma.ncirl.ie/id/eprint/7757

Actions (login required)

View Item