Application Of Data Mining For Prediction Of Sales Of Best-Selling Electronic Products Using K-Nearest Neighbor Method

Yulia Rizki Amalia (Universitas Islam Negeri Raden Fatah Palembang)

Jurnal Mandiri IT
Vol. 8 No. 2 (2020): January: Coputer Science and related.

PT. Bintang Multi Sarana Palembang is one of the largest electronics companies in South Sumatra. This company has a wide variety of electronic products on offer. Judging from the large number of consumer demands for electronic products based on sales data for the last 3 years, predictions are needed for sales of the best-selling electronic products, in order to make it easier for the company to plan stock supply. To determine the sales of the best-selling electronic products, data mining classification techniques and the K-Nearest Neighbor algorithm are used. The results of this study are predictions of the best selling electronics sales as many as 6 types of products from 22 types of products sold, namely CTV, Refrigerator, DVD, Speaker, Washing Machine and LCD. Based on the accuracy value of the best-selling product sales classification of 92.51%.