Predictive Analytics For Stroke Disease

Hospitals take possession of daily large amounts of data of their patients who suffer from various illnesses, including stroke, and the lack of analytical tools used to obtain information in that regard, has made it difficult for physicians and paramedics to gain knowledge. In order to address this issue, predictive analytics serves to predict the type of stroke disease beforehand hence the treatment actions can be carried out early and more appropriately to avoid worsening the patients' condition status. In this research, we presented a model for predicting stroke solely based on the patients' medical record. The model was built by using artificial neural network to produce an estimation of the type of stroke whether it gets any worse than the initial diagnosis. Prediction analysis using Artificial Neural Network method possessed accuracy of 95.15%.