Identifying the Patients at Risk of Stroke Using Anomaly Detection Based Classification Approach

Jagwani, Girish (2019) Identifying the Patients at Risk of Stroke Using Anomaly Detection Based Classification Approach. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (832kB) | Preview
Preview	PDF (Configuration manual) Download (1MB) | Preview

Abstract

Stroke is one of the leading causes for death in 21st Century, accounting for the death of more than 2000 people in Ireland every year. Health Care Industry has done a lot of progress to cure stroke, but Stroke strikes suddenly, and the damage rate is so high that even if cured, it leaves permanent disabilities. The aim of this project is to identify the patients at risk of stroke using Electronic Health Records available with the hospitals and medical institutions. This is achieved by developing an Ensemble Voting Classifier with 9 different classification models as predictors. As the healthcare datasets are prone to be highly imbalanced, the 9 classification models along with the Ensemble Voting Classifier are developed and evaluated using 3 different sampling techniques. While evaluating the performance of all 30 modelled combinations, the combination of Ensemble Voting Classifier and hybrid sampling technique (SMOTE + Tomek) achieved the best results. The results obtained are promising and have successfully contributed towards the stroke detection problem in the healthcare industry.

Item Type:	Thesis (Masters)
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QA Mathematics > Computer software T Technology > T Technology (General) > Information Technology > Computer software R Medicine > R Medicine (General)
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Dan English
Date Deposited:	16 Jun 2020 10:05
Last Modified:	16 Jun 2020 10:05
URI:	https://norma.ncirl.ie/id/eprint/4289

Actions (login required)

View Item