Comparison of Linear Regression, Neural Net, and Arima Methods For Sales Prediction of Instrumentation and Control Products In PT. Sarana Instrument

Abstract

PT. Sarana Instrument is a national private company which is an authorized sales agent for several instrumentation and control products originating from European countries and the United States for sales in Indonesia. Every company certainly targets sales to be achieved every year, for that the company certainly needs sales forecasting. PT. Sarana Instrument does not currently have a prediction system so that for making annual sales targets, it still uses manual estimates by looking at sales data from the previous year's sales. So that PT. Sarana Instrument cannot get accurate sales predictions and the company cannot prepare human resources and financial resources according to the company's needs. Therefore, a forecasting system is needed to help make forecasts. The purpose of this study was to analyze the error rate of forecasting sales data for 2013-2021 at PT. Sarana Instrument uses a forecasting algorithm, namely the Linear Regression Algorithm, Neural Net and Arima, in order to obtain a sales forecast method with the smallest error rate and can be implemented at PT. PT. Sarana Instrument.