Penerapan Algoritma Regresi Linear untuk Prediksi Penjualan Beberapa Produk Makanan dan Minuman (Studi Kasus: Kantin Berkat)

Abstract

Sales are activities that provides benefits for business operators, including UMKM. Every business operator has documentation of sales data. Kantin Berkat, UMKM which sells food and beverages, always documents their daily sales. Documentation is to record the daily sales of products. Daily sales of products in Kantin Berkat vary that some products show sales fluctuations. The impact is some products must be wasted because they don’t sell on certain days and must have affected the gain at Kantin Berkat. This research attempts to predict the sales amount for the next days based on sales data of two years prior. The prediction’s done with the Simple Linear Regression Algorithm by Python encoding for five products, like seblak, dimsum, singkong, es jeruk, and es teh manis. The results of testing the algorithm showed that the MAPE values for the prediction results of the five products were 35,86%, 28,17%, 48,20%, 15,92%, and 20,03%.