Forecasting Agricultural Seeds Sales System Using The Additive Holt-Winters Method

¹ Informatics Department, University of Jember, Indonesia
² Information System Department, University of Jember, Indonesia
³ Information Technology Department, University of Jember, Indonesia

^* Corresponding author: qurrotaaar@unej.ac.id

Abstract

Buana Seed is an agricultural business sector that sells various kinds of agricultural seeds, such as: chilli seeds, Servo tomato seeds, and Ranti tomato seeds. The problem found was the sale of seeds that was not optimal, because the owner used the previous period's sales as a reference. Due to an error in determining the number of seeds to be sown, there is often an overstock or understock. An information system is needed to assist seeders in determining the number of seeds to be sown. The forecasting method used is the "Additive Holt-Winters" method where the data used is sales data for each seed from January 2019 to August 2021. This method is based on three elements, namely linear, trend and seasonal data elements by giving three weights respectively in their predictions, namely α, β, and γ. There were 10 combinations of alpha (α), beta (β), and gamma (γ) made in the calculation. The accuracy value will be calculated using MAPE. MAPE results from each combination are different. The results of this forecasting calculation will be a reference in determining the seeds to be sown so that sales become optimal.