Prediksi menggunakan Metode Extreme Learning Machine untuk Penjualan Produk Kopi Bubuk (Studi Kasus Kafe Tjap Daoen Bondowoso)

Abstract

Cafe Tjap Daoen, has not used a method to support sales forecasting so that the cafe only determines the sales of the cafe owner with reference to the previous period without any calculations resulting in underproduction when busy, overproduction of coffee when there are several orders. This study uses two Arabica Specialty Coffee and Arabica Wine products with data for 2019 – 2022 respectively. The Extreme Learning Machine (ELM) algorithm has very good performance in predicting time series data. The results of the research conducted, the ELM Algorithm is able to produce MAPE with a percentage of 1.2658% and MAD 3.1784 Arabica Specialty Coffee, Arabica Wine MAPE 10.373% MAD 3.5097.