Forecast of the Trend in Sales Data of a Confectionery Baking Industry Using Exponential Smoothing and Moving Average Models

Rasaq A. Kazeem*, Moses O. Petinrin, Peter O. Akhigbe, Tien-Chien Jen, Esther Akinlabi, Stephen A. Akinlabi, Omolayo M. Ikumapayi

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Starch-containing foods such as bread, pastries, and cakes are usually baked at a moderately high temperature in an oven. When these products are later exposed to room temperature, the associated gelatinized starch begins to harden which causes retrogradation and molecular realignment. Due to this circumstance, manufacturers need to have a fairly accurate estimate of products demand in order to determine the precise amount of baking powder and additives for use in their production so as not to incur losses in their business arising from the stale and consequentially unsalable products. This research was therefore focused on selecting the best forecasting model using a prominent confectionery firm in Abeokuta, Ogun State, Nigeria as a case study. The study was based on 24-week operational period sales data collected from the company. The moving average model and the exponential smoothing model were the two forecasting models considered in this research. The data obtained was thoroughly reviewed and the results of the forecasting models were compared. The most effective model was the exponential smoothing model as it produced the lowest mean absolute percentage error on the average of 3.7347 for the cumulative days of sales under review as against the 15.1713 for the moving average model. However, the exponential smoothing model was considered the best forecasting model for minimizing forecasting error in this study.
Original languageEnglish
JournalMathematical Modelling of Engineering Problems
Publication statusAccepted/In press - 3 Oct 2022

Fingerprint

Dive into the research topics of 'Forecast of the Trend in Sales Data of a Confectionery Baking Industry Using Exponential Smoothing and Moving Average Models'. Together they form a unique fingerprint.

Cite this