Prediksi Penjualan Produk Rokok Pada PT. Indomarco Prismatama Menggunakan Algoritma C4.5

Abstract

- The Covid-19 pandemic has had a bad influence on the national economy throughout 2020, the impact of which can be felt, including by retail. The purpose of this study is to inform the prediction of sales of cigarette products at Indomaret Plus Eco RSUD Cengkareng so that companies can procure goods according to demand. The correct prediction is expected to overcome several problems in selling cigarette products at Indomaret Plus Eco RSUD Cengkareng, such as piling up and product irregularities, making the store have to continue to carry out FEFO (First Expired First Out) which in practice is quite a time consuming or overproduction and under production. The method in this research is the C4.5 Algorithm. The C4.5 algorithm is a data classification algorithm with the type of decision tree (decision tree). The highest accuracy value was obtained from experiments with 58 training data and 37 testing data, which showed an accuracy rate of 92.11%, an error rate of 7.894%, and an AUC value of 87.8% with good classification accuracy. Size is the attribute that most influences the sale of cigarette products, which are classified as selling or not selling well.