Implementasi Metode Fuzzy Time Series dalam Peramalan Penjualan Produk Unggulan Perusahaan

Abstract

This research discusses sales forecasting of two superior products of PT. Pakis Logam Perkasa Indonesia: semi-stainless spatula and electric elbow grater c. One of the household furniture manufacturers in Tulungagung sells its wholesale and retail products. Due to the high demand for semi-stainless spatula and c-elbow electric grater products, the company experienced uncontrolled inbound and outbound sales. Therefore, sales forecasting is imperative. Sales forecasting uses previous data to decide how many products to market. Chen's Fuzzy Time Series model uses a technique for forecasting the demand for semi-stainless spatulas and c-elbow electric grates for the next month based on previous data patterns. The two methods used to analyze data are qualitative analysis and quantitative analysis. The aim of creating this forecasting model is to improve forecasting results, especially in terms of accuracy. Next, the Mean Absolute Percentage Error (MAPE) determines the error value from the forecasting results. The error value for the semi-stainless spatula product is 1.03%, and the c-elbow electric grater is 0.62%. The MAPE test shows that the forecasting results of these two products are excellent.