Increase Accuracy of Naïve Bayes Classifier Algorithm with K-Means Clustering for Prediction of Potential Blood Donors

Abstract

Branch of Computer Science knowledge is data mining. Data mining help people to processing a big and irregular data. In public health, data mining can be used to manage blood donors data. Blood donors is a proses to take some blood from volunteer then given to other people who need. One of the ways to fill up blood requirement in Indonesia is organize blood donors event regularly, but some people didn’t routine give they blood. Solution of that problems, a system needed to predict future blood donor behavior. Recency, Frequency, Monetary, Time, Churn Probability (RFMTC) is a modification from Recency of purchase, Frequency of purchase, and Monetary value of purchase (RFM) that used to predict a blood donors behavior. In this research, implemented a Naïve Bayes Classifier to blood donors classification. The classification result with 224 data from RFMTC dataset is 78.13% accuracy. Combination Naïve Bayes Classifier algorithm with K-Means Clustering increase accuracy to 80.80%.