Penerapan Regresi Time Series, Arimax Dan Ffnn Untuk Peramalan Arus Uang Kartal Di Kantor Perwakilan Bank Indonesia Wilayah Sulawesi

Masyitha, Aulia (2015) Penerapan Regresi Time Series, Arimax Dan Ffnn Untuk Peramalan Arus Uang Kartal Di Kantor Perwakilan Bank Indonesia Wilayah Sulawesi. Undergraduate thesis, Institut Teknology Sepuluh Nopember.

[img]
Preview
Text
1311100085-Undergraduate Thesis.pdf

Download (3MB) | Preview

Abstract

Bank Indonesia dalam melaksanakan tugas mengatur dan menjaga kelancaran sistem pembayaran berwewenang dalam pengelolaan uang kartal yang meliputi perencanaan, pengeluaran (outflow), pengedaran, pencabutan dan penarikan (inflow) uang kartal. Outflow dan inflow uang kartal ada kecenderungan dipengaruhi oleh efek variasi kalender, sehingga metode peramalan yang dapat digunakan adalah regresi time series dan ARIMAX dengan efek variasi kalender. Data outflow dan inflow ada kemungkinan memiliki pola non linier sehingga metode permalan yang dapat digunakan adalah FFNN. Kriteria pemilihan model terbaik didasarkan pada RMSE out sample. Tujuan dari penelitian ini adalah meramalkan outflow dan inflow uang kartal di Kantor perwakilan Bank Indonesia wilayah Sulawesi menggunakan metode ARIMA, regresi time series, ARIMAX, dan FFNN. Berdasarkan kriteria RMSE out sample diperoleh model ARIMA sebagi model terbaik untuk meramalkan outflow uang kartal di Sulawesi Selatan dan Inflow uang kartal di Sulawesi Tengah, model regresi time series merupakan model terbaik untuk meramalkan outflow uang kartal di Sulawesi Tengah, model FFNN merupakan model terbaik untuk inflow uang kartal di Sulawesi Selatan serta outflow dan inflow uang kartal di Sulawesi Utara dan Sulawesi Tenggara ====================================================================================================== Bank Indonesia in implementing the tasks of regulating and safeguarding the smooth and efficient operation of the payment systems, authorized in the management of currency which includes planning, outflow, distribution, revocation and inflow. Outflow and inflow of currency there is a trend influenced by the effects of calendar variations, so the forecasting methods that can be used is a time series regression and ARIMAX with calendar variations effect. Data outflow and inflow is likely to have a non linear pattern so that the forecasting methods that can be used is FFNN. Criteria for selecting the best model based on the RMSE out sample. The purpose of this research is to predict the outflow and inflow of currency in Bank Indonesia representative office Sulawesi using ARIMA, regression time series, ARIMAX, and FFNN. Based on the criteria RMSE out sample obtained ARIMA model as a best model to predict the outflow of currency in South Sulawesi and Inflow currency in Central Sulawesi, time series regression model is the best model to predict the outflow of currency in Central Sulawesi, FFNN the model is the best model for the inflow currency in South Sulawesi as well as the outflow and inflow of currency in North Sulawesi and Southeast Sulawesi.

Item Type: Thesis (Undergraduate)
Additional Information: RSSt 519.536 Mas p
Uncontrolled Keywords: ARIMAX, FFNN, Inflow, Outflow, Regresi Time Series, RMSE, Uang kartal.
Subjects: Q Science > QA Mathematics > QA278.2 Regression Analysis
Divisions: Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User: Mr. Tondo Indra Nyata
Date Deposited: 08 Nov 2019 03:17
Last Modified: 08 Nov 2019 03:17
URI: https://repository.its.ac.id/id/eprint/71659

Actions (login required)

View Item View Item