Estimasi Parameter Dan Pengujian Hipotesis Geographically Weighted Bivariate Zero Inflated Poisson Inverse Gaussian Regression Studi Kasus : Jumlah Kasus Kusta Basah Dan Kusta Kering Di Kabupaten Banjarnegara Dan Kabupaten Kebumen Tahun 2018

Nur, Muhammad Saifudin (2022) Estimasi Parameter Dan Pengujian Hipotesis Geographically Weighted Bivariate Zero Inflated Poisson Inverse Gaussian Regression Studi Kasus : Jumlah Kasus Kusta Basah Dan Kusta Kering Di Kabupaten Banjarnegara Dan Kabupaten Kebumen Tahun 2018. Masters thesis, Institut Teknologi Sepuluh Nopember.

[img] Text
06211850012008-Master_Thesis.pdf
Restricted to Repository staff only

Download (4MB) | Request a copy
[img] Text
06211850012008-Master_Thesis.pdf
Restricted to Repository staff only

Download (4MB) | Request a copy
[img] Text
06211850012008-Master_Thesis.pdf - Accepted Version
Restricted to Repository staff only until 1 April 2024.

Download (4MB) | Request a copy

Abstract

Zero Inflated Poisson Inverse Gaussian Regression (ZIPIGR) merupakan pengembangan regresi yang mampu mengakomodasi under/overdispersi dan banyaknya nilai nol pada data observasi. ZIPIGR bersifat model regresi yang global, sehingga intepretasi model yang diperoleh berlaku sama pada seluruh observasi. ZIPIGR yang memperhatikan faktor lokasi yakni koordinat lintang dan bujur yang melekat pada data observasi saat dua variabel respon saling berkorelasi yaitu model GWBZIPIGR. Pada penelitian ini akan dibahas estimasi parameter menggunakan Maximum Likelihood Estimation (MLE), serta digunakan iterasi numerik Berndt Hall Hall Hausman (BHHH) apabila persamaan yang dihasilkan tidak close form dan pengujian hipotesis untuk model GWBZIPIGR menggunakan Maximum Likelihood Ratio Test (MLRT). Pemodelan regresi model Poisson GWBZIPIGR jumlah kasus kusta kering menghasilkan hampir variabel signifikan pada setiap kecamatan kecuali pada 7 kecamatan yang tidak signifikan pada variabel kepadatan penduduk. Pemodelan model poisson GWBZIPIGR jumlah kasus kusta basah menghasilkan hampir variabel signifikan pada setiap kecamatan kecuali pada 8 kecamatan yang tidak signifikan pada variabel kepadatan penduduk. Model GWBZIPIGR dinilai lebih baik dibandingkan model BZIPIGR karena memiliki nilai SSE yang lebih kecil. ===================================================================================================== Zero Inflated Poisson Inverse Gaussian Regression (ZIPIGR) is a regression development that is able to accommodate under/overdispersion and the number of zero values in the observation data. ZIPIGR is a global regression model, so the interpretation of the model obtained is the same for all observations. ZIPIGR which pays attention to the location factor, namely the latitude and longitude coordinates attached to the observation data when the two response variables are correlated with each other, namely the GWBZIPIGR model. In this study, parameter estimation using Maximum Likelihood Estimation (MLE) will be discussed, and Berndt Hall Hall Hausman (BHHH) numerical iteration will be used if the resulting equation is not in close form and hypothesis testing for the GWBZIPIGR model using the Maximum Likelihood Ratio Test (MLRT). Poisson GWBZIPIGR regression modeling the number of dry leprosy cases resulted in almost a significant variable in each sub-district except for 7 sub-districts which were not significant in the population density variable. The Poisson GWBZIPIGR modeling the number of wet leprosy cases resulted in almost significant variables in each sub-district except for 8 sub-districts which were not significant in the population density variable. The GWBZIPIGR model is considered better than the BZIPIGR model because it has a smaller SSE value.

Item Type: Thesis (Masters)
Uncontrolled Keywords: GWBZIPIGR, MLE, MLRT, Jumlah Kasus Kusta. Number of Leprosy
Subjects: H Social Sciences > HA Statistics
H Social Sciences > HA Statistics > HA29 Theory and method of social science statistics
Q Science > QA Mathematics > QA278 Cluster Analysis. Multivariate analysis. Correspondence analysis (Statistics)
Divisions: Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49101-(S2) Master Thesis
Depositing User: Muhammad Saifudin Nur
Date Deposited: 23 Feb 2022 20:55
Last Modified: 23 Feb 2022 20:55
URI: https://repository.its.ac.id/id/eprint/94733

Actions (login required)

View Item View Item