Page 6 - FULL REPORT 30012024

ABSTRACT

Addressing key challenges in stroke risk assessment within healthcare, such as the lack

of customization, accessibility, and efficiency, this project developed a web
application utilizing Random Forest algorithm. This study involved an analysis of

existing stroke risk assessment techniques and the creation of a tailored machine
learning tool, followed by the prediction model evaluation and extensive user

acceptance testing with a diverse group of participants. The web application features
a dashboard that displays stroke mortality data and includes a stroke risk prediction

tool enhanced by machine learning. User feedback, gathered through comprehensive

testing, shows positive result, underscoring the application's role in raising stroke risk
awareness and its ease of use. The prediction model's evaluation further strengthens

the application's credibility by demonstrating its effectiveness in correctly identifying
stroke cases. It is recommended that further research be undertaken in expanding

collaborative efforts with healthcare and research institutions to enrich data sources
and insights, integrating real-time health metrics to increase the tool's relevance, and

improving user interaction and engagement to enhance the application's appeal and

usability. These steps are vital for extending the application's reach and inclusivity,
thereby maximizing its effectiveness in diverse healthcare environments.

iv