Sales Forecast Optimization: Ensemble and Time Series Comparison

Abstract

To ensure competitive advantage, companies seek solutions that allow them to optimize the management of their resources.   Sales forecasting is the process of organizing and analyzing information in a way that allows estimating how sales will be.  In this context, decision support systems can be allies to explore scenarios based on historical data. This is an essential and inexpensive way for every company to increase its profits, decrease its costs, and achieve greater flexibility to change. The objective of this study is to analyze a sales database to predict the amount of sales. Thus, this research aims to compare the Ensemble and Time Series (ARIMA) forecasting methods in order to find the most optimized model for the proposed problem. The preliminary results of this study showed inconclusive results with 5% significance that there was a change in performance between the two approaches.