Perbandingan Algoritma Klasifikasi Data Mining Untuk Prediksi Penyakit Kulit

Abstract

Skin diseases often thought as a insignificant problem compared to diseases with high mortality, such as HIV/AIDS, ischemic heart disease, stroke, tuberculosis, malaria and cancer. However, skin diseases are some of the most common diseases seen in the developing country. Skin diseases diagnosis and treatment is important because there are a lot of skin diseases that shows similar symptoms and screening for sign of skin disease is an important way for systemic diseases. Prediction of skin disease is diffucult because of a lot of skin disease shows similar symptoms. Data mining with clasification algorithm can be used to predict skin disease. Four data mining clasification algorithm C4.5, Naive Bayes, KNN and SVM was used for data analysis. Medical dataset used in this research is dataset download from UCI repository site. From the tes result by comparing accuration, precision dan recall, it is known that Naive Bayes has the highest accuracy and precision.