Klasifikasi Risiko Penyakit Paru Menggunakan Algoritma Decision Tree Menggunakan RapidMiner

Abstract

Lung disease is one of the health conditions that can be influenced by various factors, such as age, smoking habits, pollution exposure, family history, and respiratory symptoms. The application of data mining techniques can assist in identifying the level of lung disease risk based on patient characteristics. This study aims to classify the risk level of lung disease using the Decision Tree algorithm. The dataset used consists of 250 patient records with attributes including age, smoking habits, chronic cough, shortness of breath, history of pneumonia, pollution exposure, family history, and lung disease risk level. The classification process was carried out using RapidMiner through several stages, including label determination, training and testing data partitioning, Decision Tree model construction, and evaluation using a Confusion Matrix. The testing results showed an accuracy value of 34.67%. The resulting decision tree indicates that age is the most influential factor in determining the risk level of lung disease, followed by pollution exposure, family history, chronic cough, and shortness of breath. Although the obtained accuracy is relatively low, this study demonstrates that classification methods can be utilized to analyze the factors affecting lung disease risk.