Solekha, Novia Amilatus (2024) Model Mixed Geographically And Temporally Weighted Bivariate Weibull Regression (Studi Kasus: Kemiskinan Dan Angka Harapan Hidup Di Provinsi Jawa Timur Tahun 2018-2022). Masters thesis, Institut Teknologi Sepuluh Nopember.
Text
6003212023-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 July 2026. Download (4MB) | Request a copy |
Abstract
Geographically Weighted Bivariate Weibull Regression (GWBWR) merupakan perkembangan dari metode regresi bivariat yang digunakan untuk memodelkan proses yang heterogen secara spasial dengan variabel respon berdistribusi Weibull. Dalam pengembangannya, terdapat banyak kasus di mana informasi dari data panel dibutuhkan. Penggunaan data panel memberikan keunggulan informasi yang lebih lengkap karena mencakup beberapa periode waktu, tetapi juga memperkenalkan kemungkinan adanya efek temporal. Model regresi spasial temporal dalam situasi tertentu dapat menunjukkan bahwa variabel prediktor memiliki pengaruh tidak hanya secara lokal, tetapi juga secara global. Untuk mengatasi kondisi tersebut, penelitian ini mengembangkan model Geographically and Temporally Weighted Bivariate Weibull Regression (GTWBWR) menjadi Mixed Geographically and Temporally Weighted Bivariate Weibull Regression (MGTWBWR). Penaksiran parameter dalam model dilakukan menggunakan metode Maximum Likelihood Estimation (MLE). Pengujian serentak dilakukan menggunakan metode Maximum Likelihood Ratio Test (MLRT), sementara pengujian parsial menggunakan uji Z. Aplikasi model MGTWBWR diterapkan pada data kemiskinan dan Angka Harapan Hidup (AHH) di Jawa Timur. Variabel prediktor yang digunakan dalam penelitian ini meliputi rata-rata lama sekolah, tingkat pengangguran terbuka, persentase keluhan kesehatan, persentase rumah tangga dengan akses sanitasi layak, dan rasio Gini. Hasil pemodelan menunjukkan bahwa faktor-faktor tersebut memiliki pengaruh yang berbeda-beda terhadap kemiskinan dan AHH di setiap lokasi di Jawa Timur. Berdasarkan faktorfaktor yang mempengaruhi kemiskinan tahun 2022 dengan menggunakan data 5 periode. Provinsi Jawa Timur dibagi menjadi enam kelompok, sedangkan untuk AHH, terbagi menjadi delapan kelompok. Model MGTWBWR yang digunakan untuk menganalisis kemiskinan dan AHH membagi Jawa Timur menjadi 10 kelompok. Selain itu, diketahui bahwa variabel yang bersifat global untuk kedua variabel respon adalah gini rasio
====================================================================================================================================
Geographically Weighted Bivariate Weibull Regression (GWBWR) is a development of the bivariate regression method which is used to model spatially heterogeneous processes with Weibull distributed response variables. In its development, there are many cases where information from panel data is needed. The use of panel data provides the advantage of more complete information because it covers several time periods, but also introduces the possibility of temporal effects. Spatial temporal regression models in certain situations can show that predictor variables have an influence not only locally, but also globally. To overcome this condition, this research developed the Geographically and Temporally Weighted Bivariate Weibull Regression (GTWBWR) model to become Mixed Geographically and Temporally Weighted Bivariate Weibull Regression (MGTWBWR). Parameter estimation in the model was carried out using the Maximum Likelihood Estimation (MLE) method. Simultaneous testing was carried out using the Maximum Likelihood Ratio Test (MLRT) method, while partial testing used the Z test. The MGTWBWR model application was applied to poverty and Life Expectancy Rate (AHH) data in East Java. Predictor variables used in this research include average years of schooling, open unemployment rate, percentage of health complaints, percentage of households with access to proper sanitation, and the Gini ratio. The modeling results show that these factors have different influences on poverty and AHH in each location in East Java. Based on factors influencing poverty in 2022 using data from 5 periods. East Java Province is divided into six groups, while for AHH, it is divided into eight groups. The MGTWBWR model used to analyze poverty and AHH divides East Java into 10 groups. Apart from that, it is known that the global variable for the two response variables is the Gini ratio
Item Type: | Thesis (Masters) |
---|---|
Uncontrolled Keywords: | GTWBWR, MGTWBWR, Kemiskinan, Angka Harapan Hidup, Poverty , Life Expectancy RateGTWBWR, MGTWBWR, Poverty , Life Expectancy Rate |
Subjects: | G Geography. Anthropology. Recreation > G Geography (General) Q Science |
Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49101-(S2) Master Thesis |
Depositing User: | Novia Amilatus Solekha |
Date Deposited: | 20 Feb 2024 06:26 |
Last Modified: | 20 Feb 2024 06:26 |
URI: | http://repository.its.ac.id/id/eprint/107614 |
Actions (login required)
View Item |