Stroke Prediction Using Machine Learning: A Review

Abstract

A stroke occurs when blood supply to the part of brain is reduced or interrupted causes blockage in an artery which is serious issue. It is second major reason for deaths in worldwide. It is caused due to people lifestyle decision, high blood sugar, heart disease, obesity, hypertension. Due to this prediction of stroke becomes necessary and with the help of effective prediction algorithm which allow for early diagnosis and intervention. Several models are developed and evaluated to design a robust framework for long term risk prediction of stroke occurrence. Using different machine learning algorithm models namely gaussian naive bayes, logistic regression, decision tree, k nearest neighbour. The efficient data collection, data preprocessing, data transformation methods applied to provide reliable information for model to be successful. The performance of each classifier is estimated based on evaluation metrics such as accuracy, error rate loss function. It has possible to obtain accuracy of 98 %.