Penentuan Spesifikasi Sofa dengan Decision Tree pada Perusahaan X

Abstract

Company X is a company engaged in the furniture sector. The company has been established since 2003. Company X has a problem, which is that it is still unable to design new product specifications and determine which current products need to be produced again using past demand data. This causes high RnD (Research and Development) cost losses. These costs include drafting costs, costs for making frame molds, fabrics, foam, and prototype costs. With the application of BI to company X, the specifications of the sofa can be found out according to the interests and needs of consumers. The data used in this study are sales from January 1, 2021 to June 30, 2022. The method used to conduct this research is the decision tree version ID3 and C4.5. After being tested using cross-validation 10 times, the results of the accuracy show that the C4.5 version of the decision tree is more accurate than the ID3 version. The results of this study are to determine the sofa specifications that customers are interested in based on price segmentation. Categories based on price are divided into 3, namely PR3, PR2, and PR1. There are 6 significant sofa specifications in PR2. There are 3 significant sofa specifications in the PR3.