Abstract—A stroke is the result of cell death caused by poor blood flow or vascular obstruction in the brain, but it normally happens suddenly and is hard to prevent. In addition, strokes are one of the main causes of death, with many people dying from this disease every year. Hence, using the latest technology to predict strokes is an important concern.
In this study, we built a classifier based on the simulation and practical data both in 50 data sets from Japanese patients. We used attribute selection (CfsSubsetEval evaluator and Greedy Stepwise method) and Decision Trees (Random Tree algorithm) to build the classifier with high accuracy. In the end, the result demonstrates that our method, using data that merges the simulation and practical data, can achieve high accuracy.
Index Terms—Stroke prediction, random tree algorithm, WEKA, select attributes, machine learning.
Y. C. Chen is now with the Tokyo University of Science, Noda City, 278-8510 Japan (e-mail: firstname.lastname@example.org).
T. Suzuki is now with the Tokyo University of Science, Katsushika-ku, 125-8585 Japan (e-mail: email@example.com).
M. Suzuki and H. Ohwada are now with the Department of Industrial Administration, Tokyo University of Science, Noda City, 278-8510 Japan (e-mail: firstname.lastname@example.org, email@example.com).
H. Takao is now with the Department of Neurosurgery and Innovation for Medical Informaion Techology, Jikei University School of Medicine, Tokyo, 105-8461 Japan (e-mail: firstname.lastname@example.org).
Y. Murayama is now with the Department of Neurosurgery, Jikei University School of Medicine, Tokyo, 105-8461 Japan (e-mail: email@example.com).
Cite: Yu-Chen Chen, Takashi Suzuki, Masaaki Suzuki, Hiroyuki Takao, Yuichi Murayama, and Hayato Ohwada, "Building a Classifier of Onset Stroke Prediction Using Random Tree Algorithm," International Journal of Machine Learning and Computing vol. 7, no. 4, pp. 61-66, 2017.