Prediction of Stroke Risk through Stacked Topology of ANN Model

Sabibullah M, Kasmir Raja S.V


Artificial Neural Networks (ANN) is a very popular type of Machine Learning (ML), suited for analyzing the medical data. Estimation
of stroke risks in population is not only helpful for healthcare providers but also important to identify persons at elevated risk and to select
proper treatments in clinical trials. More individual risk factors may help to improve the individual risk assessment. The objective of this study is
to predict the stroke risk by proposing the stacked ANN topology model with higher prediction accuracy. The proposed model is tested by using
three sets of real stroke population data (300 samples) and validated through statistical metrics. Our model achieved 95.33% and 94% of
accuracy in training and testing phase respectively. The obtained experimental results predicted that it is a high rate of correctness in the stroke
risk prediction task.


Keywords: Stroke Risk, Artificial Neural Networks, Risk Factors, Stacked Topology, Machine Learning, Prediction.

Full Text:




  • There are currently no refbacks.

Copyright (c) 2016 International Journal of Advanced Research in Computer Science