PENERAPAN DATA MINING UNTUK PREDIKSI MEREK PAKAIAN YANG PALING DIMINATI DENGAN METODE K-NEAREST NEIGHBOR (STUDI KASUS : PT. MATAHARI DEPARTEMENT STORE BINJAI)

Abstract

One of the business activities that must be carried out to keep the company running and growing is sales. Decisions taken by corporate responsibility holders will affect the company in the future. One of the decisions that must be determined is the product to be sold for the next period. In determining the decision, a method is needed so that the decisions to be taken can be right on target. The technique used to predict the situation in the next period is called prediction. This study proposes the development of a clothing sales prediction application. The method used is the classification with the K-Nearest Neighbor algorithm. The results of data mining calculations using classification techniques with the K-Nearest Neighbor algorithm are the most predominant, it can be predicted that the number of clothing sales in the next period will increase with an average prediction of 14,900 per month and the most popular clothing brand is Cardinal.