Issue |
ITM Web Conf.
Volume 70, 2025
2024 2nd International Conference on Data Science, Advanced Algorithm and Intelligent Computing (DAI 2024)
|
|
---|---|---|
Article Number | 04029 | |
Number of page(s) | 10 | |
Section | AI and Advanced Applications | |
DOI | https://doi.org/10.1051/itmconf/20257004029 | |
Published online | 23 January 2025 |
Stroke Prediction Based on Machine Learning
Marlan and Rosemary Bourns College of Engineering, University of California, Riverside, 92521, the United States
Corresponding author: yzhan1111@ucr.edu
Stroke has become an important cause of death and disability worldwide, which highlights the need for early detection and intervention. Machine learning technology can analyze patients’ historical health data and biometrics to identify high-risk individuals in a timely manner, thereby effectively predicting stroke.This paper evaluates the predictive performance Random Forest and Support Vector Machine (SVM). Data preprocessing encompasses managing missing data, processing categorical variables, and tackling issues related to class imbalance. Analysis of the quantitative results indicates that the Random Forest model reaches an accuracy of 95% and a precision of 93%, providing a slight edge over the SVM, which records an accuracy of 92% and a precision of 90%.. However, both models exhibit high false-negative rates, with Random Forest showing a false-negative rate of 12% and SVM at 15%, which significantly impacts their clinical utility. To improve performance, further model optimization, such as adjusting class weights or employing ensemble methods, is necessary to reduce these false-negative rates and enhance diagnostic accuracy. This study highlights the potential and limitations of machine learning in stroke prediction, showing that people need further optimization to enhance diagnostic performance.
© The Authors, published by EDP Sciences, 2025
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.