Jurnal Eurekamatika
Vol 5, No 1 (2017): Jurnal EurekaMatika

APLIKASI MULTIVARIATE GEOGRAPHICALLY WEIGHTED REGRESSION MENGGUNAKAN SOFTWARE MATLAB

Andini, Puty (Unknown)
Herrhyanto, Nar (Unknown)
Suherman, Maman (Unknown)



Article Info

Publish Date
08 Mar 2018

Abstract

ABSTRAK: Persamaan regresi dapat diterapkan pada kasus derajat kesehatan masyarakat. Indikator derajat kesehatan yang dalam penulisan ini adalah angka harapan hidup dan persentase gizi buruk balita. Jika persamaan mengandung unsur spasial maka tidak bisa di selesaikan menggunakan regresi global karena akan menyebabkan kesimpulan yang diambil kurang tepat. Keragaman spasial akibat faktor geografis dapat diselasaikan dengan pendekatan titik. Pendekatan titik yang digunakan dalam penulisan ini adalah Multivariate Geographically Weighted Regression (MGWR). Menggunakan dua variabel respon, yaitu angka harapan hidup (Y1) dan persentase gizi buruk balita (Y2) dan enam variabel prediktor, yaitu angka melek huruf (X1), rata-rata lama sekolah (X2) Indeks kesejahteraan rakyat (X3), persentase Penduduk dengan sumber air minum PDAM/pipa (X4), persentase penduduk dengan jarak sumber air minum ke tempat penampungan kotoran 10 meter (X5), dan pengeluaran perkapita penduduk untuk makan (X6). Dalam model MGWR penaksiran parameter diperoleh dengan menggunakan Weighted Least Square (WLS). Fungsi pembobotan yang digunakan fixed kernel Gaussian. Penelitian ini membuat model  derajat kesehatan dengan metode MGWR menggunakan software Matlab dan membuat peta sebaran kabupaten dan kota di Jawa Barat berdasarkan variabel-variabel yang signifikan terhadap derajat kesehatan. Variabel yang di duga mempengaruhi angka harapan hidup (Y1) adalah rata-rata lama sekolah (X2) dan variabel yang di duga mempengaruhi persentase gizi buruk balita (Y2) adalah Angka Melek Huruf (X1). Kata Kunci: Derajat Kesehatan, Multivariate Geographically Weighted Regression (MGWR)   ABSTRACT: The regression equation can be applied in cases of public health degree. Indicators of health, which in this paper are the life expectancy and the percentage of malnourished children under five years old. If the equation contains spatial element it can not be resolved using the global regression because it would lead to inaccurate conclusions drawn. Spatial heterogeneity due to geographical factors can be solved approach point. Point approach used in this paper is Multivariate Geographically Weighted Regression (MGWR) with two response variables, namely life expectancy (Y1) and the percentage of malnourished children under five (Y2) and six predictor variables, ie the literacy rate (X1), the mean the duration of schooling (X2) Index welfare of the people (X3), the percentage of residents with drinking water taps / plumbing (X4), the percentage of people with a distance of drinking water source to the septic tank of 10 meters (X5), and spending per capita for the population to eat (X6). In the model parameter estimation MGWR obtained by using Weighted Least Square (WLS). Weighting function used fixed kernel Gaussian. This study makes a model using methods MGWR health status with Matlab Software and create distribution maps districts / cities in West Java based on the variables that significantly influence health status. Variables are assumed to affect life expectancy (Y1) is the average length of the school (X2) and the variables that presumably affect the percentage of malnourished children under five (Y2) is ie the literacy rate (X1). Keywords: Health Status, Multivariate Geographically Weighted Regression (MGWR)

Copyrights © 2017






Journal Info

Abbrev

JEM

Publisher

Subject

Computer Science & IT Industrial & Manufacturing Engineering Mathematics

Description

Jurnal EurekaMatika (e-ISSN: 2528-4231, p-ISSN: 2776-480X) was first published annually on December 2013, and then since 2017 has been published twice a year, on May and November. JEM is a peer-reviewed Mathematics journal with its scope covers Algebra, Analysis, Statistics, and Applied Mathematics. ...