Brain stroke prediction using stacked ensemble model

Stroke is a potentially fatal illness that requires emergency care. There is a greater chance that the patient will recover and resume their regular life when they receive treatment and diagnosis as soon as feasible. Artificial Intelligence has the potential to significantly impact stroke diagnosis...

Full description

Bibliographic Details
Main Authors: Hemalatha Gunasekaran, Angelin Gladys, Deepa Kanmani, Rex Macedo, Wilfred Blessing N R
Format: Article
Language:English
Published: Penerbit Universiti Kebangsaan Malaysia 2024
Online Access:http://journalarticle.ukm.my/25591/
http://journalarticle.ukm.my/25591/1/kejut_38.pdf
Description
Summary:Stroke is a potentially fatal illness that requires emergency care. There is a greater chance that the patient will recover and resume their regular life when they receive treatment and diagnosis as soon as feasible. Artificial Intelligence has the potential to significantly impact stroke diagnosis and facilitate prompt patient treatment for physicians. Machine learning can be utilized in stroke prediction by evaluating huge volumes of patient data and detecting patterns and risk variables that may contribute to the likelihood of a stroke. In this study, we explored a stacked ensemble model that uses four base models—Decision Tree, XGBoost, RandomForest, and ExtraTree classifiers to predict the stroke. We discovered that the accuracy of the stacked ensemble model was 96.35%, higher than that of the traditional machine-learning models, other ensemble models, and ANN model.