Perbandingan Model Hybrid ARIMAX-FFNN-EGARCH dan Model Hybrid SETAR-EGARH Untuk Peramalan (Studi Kasus: Data Cash Outflow dan Inflow Bank Indonesia Kota Kediri)

Monica, Marieta (2022) Perbandingan Model Hybrid ARIMAX-FFNN-EGARCH dan Model Hybrid SETAR-EGARH Untuk Peramalan (Studi Kasus: Data Cash Outflow dan Inflow Bank Indonesia Kota Kediri). Masters thesis, Institut Teknologi Sepuluh Nopember.

	Text 6003201006-Master_Thesis.pdf Restricted to Repository staff only Download (4MB) | Request a copy
	Text 6003201006-Master_Thesis.pdf Restricted to Repository staff only Download (4MB) | Request a copy
	Text 6003201006-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2024. Download (4MB) | Request a copy

Abstract

Dalam kehidupan sehari-hari, perekonomian tak lepas dari kebutuhan akan uang. Terkait hal tersebut, dibutuhkan perencanaan pencetakan uang serta komposisi uang yang akan dicetak selama satu tahun kedepan oleh Bank Indonesia. Peramalan cash outflow dan inflow dapat digunakan untuk mengestimasikan kebutuhan uang masyarakat. Pada umumnya sering dijumpai permasalahan data deret waktu yang memiliki hubungan linier. Akan tetapi, terdapat pula data deret waktu dengan pola non-linier terutama pada bidang ekonomi. Kejadian tertentu atau terjadinya shock-shock yang menyebabkan adanya pola non-linier dan volatilitas pada data tersebut. Pemodelan non-linier yang digunakan dalam penelitian ini adalah model hybrid ARIMAX-FFNN-EGARCH dan hybrid SETAR-EGARCH. Kedua model diaplikasikan dan dibandingkan pada studi kasus data cash outflow dan inflow bulanan Kantor Perwakilan Bank Indonesia Kota Kediri. Hasil yang didapatkan yaitu penduga parameter Self-Exciting Threshold Autoregressive (SETAR) dengan metode pendugaan parameter Ordinary Least Square (OLS) terbukti memiliki sifat yang tidak bias, linier, dan memiliki varians minimum atau dapat dikatakan memenuhi sifat BLUE (Best Linear Unbiased Estimator). Model untuk peramalan data outflow dan inflow dengan kedua model dapat menangkap efek variasi kalender pola non-linier serta volatilitas yang tidak konstan. Pemodelan untuk peramalan di masa yang akan datang dapat menjadi pertimbangan penting bagi instansi terkait dalam mengambil kebijakan moneter selanjutnya.
===============================================================================================================================
In everyday life, the economy cannot be separated from the need for money. Related to this, it is necessary to plan for the printing of money and the composition of the money to be printed over the next year by Bank Indonesia. Forecasting cash outflow and inflow can be used to estimate people's money needs. In general, time series data problems are often encountered which have a linear relationship. However, there are also time series data with non-linear patterns, specially in the economic sector. Certain events or the occurrence of shocks cause non-linear patterns and volatility in the data. The non-linear modeling used in this research is the hybrid ARIMAX-FFNN-EGARCH and hybrid SETAR-EGARCH models. Both models were applied and compared to a case study of monthly cash outflow and inflow data at the Kediri City Representative Office of Bank Indonesia. The results obtained are the Self-Exciting Threshold Autoregressive estimator parameter with the Ordinary Least Square (OLS) parameter estimation method proven to have unbiased, linear properties, and has a minimum variance or can be said to meet the BLUE (Best Linear Unbiased Estimator) property. The model for forecasting outflow and inflow data with both models can capture the effects of non-linear pattern calendar variations and non-constant volatility. Modeling for future forecasts can be an
important consideration for relevant agencies in the next monetary policy.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Cash outflow-inflow, Time Series, Hybrid ARIMAX-FFNN-EGARCH, Hybrid SETAR-EGARH, Cash outflow-inflow, Deret Waktu, Hybrid ARIMAX-FFNN-EGARCH, Hybrid SETAR-EGARH
Subjects:	Q Science > QA Mathematics > QA276 Mathematical statistics. Time-series analysis. Failure time data analysis. Survival analysis (Biometry)
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49101-(S2) Master Thesis
Depositing User:	Marieta Monica
Date Deposited:	19 Feb 2022 05:43
Last Modified:	09 Mar 2023 04:09
URI:	http://repository.its.ac.id/id/eprint/94649

Actions (login required)

View Item