Perbandingan Metode Seasonal Autoregressive Integrated Moving Average (SARIMA) dan Extreme Learning Machine (ELM) pada Peramalan Peredaran Uang Kartal di Indonesia

<p>Money is generally accepted as legal tender in fulfilling an obligation. Money circulation is very important to be considered and controlled, to have a positive impact on the economy. Control of money circulation is usually emphasized on the type of cash, which is in the form of paper and metals. One of the ways that can help in controlling cash is by forecasting. This study aims to compare the accuracy of forecasting results on cash circulation data using the SARIMA and ELM methods. The data used is the circulation cash from January 2011 to April 2022. The SARIMA method is a method for forecasting time series data containing seasonality, while the ELM method is a method on artificial neural networks that can do forecasting. The best SARIMA model obtained is SARIMA (1,1,0)(0,1,0)12. The best ELM architecture obtains 12 input layer neurons, 45 hidden layer neurons, and 1 output layer. The measure of forecasting error to determine the best model is using MAPE. The results show that the SARIMA method has a training data MAPE of 2,3270% and testing data of 2,2772%, while the ELM method has a training data MAPE of 4,2548% and testing data of 3,8615%. Therefore, the SARIMA method is better than the ELM method at forecasting the circulation of cash in Indonesia.</p><p><strong>Keywords</strong>: cash; extreme learning machine; seasonal autoregressive integrated moving average.</p><p class="MsoNormal" style="margin-top: 0in; margin-right: .75pt; margin-bottom: 0in; margin-left: 0in; mso-add-space: auto; mso-para-margin-top: 0in; mso-para-margin-right: .07gd; mso-para-margin-bottom: 0in; mso-para-margin-left: 0in; line-height: normal;"> </p><table cellspacing="0" cellpadding="0" align="left"><tbody><tr><td> </td><td><img src="file:///C:/Users/lenovo/AppData/Local/Temp/msohtmlclip1/01/clip_image001.gif" alt="" width="579" height="1" /></td></tr></tbody></table><p class="MsoNormal" style="margin-top: 0in; margin-right: .75pt; margin-bottom: 0in; margin-left: 0in; mso-para-margin-top: 0in; mso-para-margin-right: .07gd; mso-para-margin-bottom: 0in; mso-para-margin-left: 0in; line-height: normal;"><!--[endif]--><em></em></p>