Institusion
Institut Teknologi Perusahaan Listrik Negara
Author
Damanik, Defri Agustian
Abdurrasyid, Abdurrasyid
Indrianto, Indrianto
Subject
Teknik Informatika
Datestamp
2022-09-26 06:19:55
Abstract :
Forecasting is a process taht will predict what will happen based on past data or information. PT Pertamina is one of the distributors of fuel oil which aims to meet the needs of the community. Currently, fuel oil stock management at gas stations has not used a forecasting method. Due to the absence of sales forecasting, the gas stations could not estimate the supply of fuel stocks for the next day. In this study a fuel oil sales forecasting system was created which later the results of the forecasting could be used by gas station officers as an estimate of fuel oil supply for the next day. This system uses the Multiple Linear Regression method which consists of the remaining independent variable (X1), the stock in (X2) and the dependent variable, the stock out (Y). After testing the classical assumptions, it can be concluded that the independent variable (X1 and X2) have an effect on the dependent variable (Y). From the results of testing the error rate using the MAPE (Mean Absolute Percent Error) method, the error rate for forecasting pertalite for a week is 11,0% and the errorr rate for forecasting diesel is 13,2%.