Penerapan Algoritma C45 Untuk Klasifikasi Keberhasilan Pengiriman Barang

Abstract

Abstrack-The success rate of shipping goods is a very important factor in the world of ecommerce business with logistics delivery services being the last link in the chain with customers. This greatly affects the satisfaction of customers who expect the goods they buy can be sent on time and in good condition when received. One technique of data mining that can be used to help predictions is to use classification techniques. Classification can be done with a decision tree that is with the C4.5 algorithm. The purpose of this study is to make a classification and apply data mining classification based on the decision tree and the rules generated. Furthermore, the results of the data classification are evaluated using a confusion matrix and ROC curve in the rapid miner application to determine the level of accuracy results. In this study produced an accuracy value of 93% and produced an AUC (Area Under Curve) value of 0.73 with a sufficient classification accuracy (Fair classification). From the results of this study can be a recommendation for the distribution section in choosing logistics services in shipping goods based on categories of goods, the destination address so that the risk of failure in shipping can be reduced.