Anggriawan, IlhamAditya (2014) Pemodelan Banyaknya Kunjungan Wisatawan Mancanegara Menggunakan Model Generalized Space Time Autoregressive Integrated-Ordinary Least Square (Gstari) Dan Seemingly Unrelated Regression (Gstari-Sur). Sarjana thesis, Universitas Brawijaya.
Abstract
Model Generalized Space Time Autoregressive Integrated (GSTARI) merupakan model deret waktu dengan melibatkan unsur lokasi. Pemodelan banyaknya kunjungan wisatawan mancanegara ke Indonesia melalui pintu masuk Bandara Internasional menggunakan model GSTARI melibatkan 4 lokasi yaitu Soekarno-Hatta, Ngurah- Rai, Kualamanu dan Juanda. Keempat lokasi ini layak dimodelkan dengan model GSTARI karena memiliki korelasi antar lokasi yang tinggi. Pemodelan GSTARI dengan pendugaan parameter Ordinary Least Square (OLS) mengasumsikan matriks ragam-peragam galat konstan. Sedangkan Seemingly Unrelated Regression (SUR) dapat mengakomodir matriks ragam-peragam galat. Model banyaknya kunjungan wisatawan mancanegara ke Indonesia menggunakan model GSTARI Ordinary Least Square (OLS) dan Seemingly Unrelated Regression (SUR) yang dihasilkan berturut-turut adalah GSTAR(21)-I(1) dan GSTAR(21)-I(1)-SUR.
English Abstract
Generalized Space Time Autoregressive Integrated (GSTARI) is a time series model including location element with non-stationary data. The modeling of Number of foreign tourists to Indonesia through International airport gate used GSTARI model consists of 4 locations which were Soekarno-Hatta, Ngurah Rai, Kualanamu, and Juanda Airport. These four models fitted to be modeled with GSTARI model because they had high correlation between location. GSTARI modeling with ordinary least square (OLS) parameter estimation produced variance-covariance matrix with constant error. While seemingly unrelated regression (SUR) could accommodate error of variance-covariance matrix. Model of number of foreign tourists to Indonesia used GSTARI model OLS and SUR were GSTAR(21)-I(1) and GSTAR(21)-I(1)-SUR. Determination coefficient of GSTAR(21)-I(1) model was 96,8% and error level (MAPE) was 6,694%. While determination coefficient of GSTAR(21)-I(1)-SUR model was 96,7% and error level (MAPE) was 7,614%. GSTAR(21)-I(1)-SUR model had close determination coefficient and error level to GSTAR(21)-I(1) model. GSTARI model with OLS-estimator produced higher determination coefficient and low error level (MAPE). This happened because GSATRI model with OLS-estimator met the assumption of non-autocorrelation between errors produced.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/MIPA/2014/239/051404653 |
| Subjects: | 500 Natural sciences and mathematics > 510 Mathematics |
| Divisions: | Fakultas Matematika dan Ilmu Pengetahuan Alam > Matematika |
| Depositing User: | Budi Wahyono Wahyono |
| Date Deposited: | 15 Aug 2014 07:35 |
| Last Modified: | 21 Oct 2021 04:21 |
| URI: | http://repository.ub.ac.id/id/eprint/153862 |
Preview |
Text
SKRIPSI_FULL.pdf Download (5MB) | Preview |
Actions (login required)
 |
View Item |