CLASSIFICATION FOR DIAGNOSING STROKE USING ORANGE DATA MINING

Abstract

Stroke is a disease with the highest mortality rate among people aged over 45 years in Indonesia. The problem with stroke in Indonesia requires very serious attention because the number of cases continues to increase, and the death rate is very high. The treatment needed is to maintain health and also detect strokes early. This research aims to build a classification model that can predict whether someone is at risk of having a stroke based on available clinical data. Factors that influence the diagnosis of stroke are needed. Based on the data obtained, several factors are used as sources of analysis such as BMI (Body Mass Index), hypertension, heart disease, glucose level, smoker or not, age, gender, type of work, and type of residence that can be used classified as input variables and output variables. Data mining can be used to classify whether a patient has had a stroke. This research aims to apply the orange data mining application using the K-Nearest Neighbour (K-NN), Naive Bayes, and Neural Network models. Next, the patient data will be analysed using the Orange Data Mining application with K-NN, Naive Bayes and Neural Network models. This research contribution can be used by health services to detect stroke early so that it is known earlier.