Leibniz: Jurnal Matematika
Vol. 6 No. 01 (2026): Leibniz: Jurnal Matematika

Comparison of Geographically Weighted Regression with Adaptive Gaussian and Bisquare Kernel on Open Unemployment Rate in Riau Islands

Widya Reza (Unknown)
Buan, Febrya Christin Handayani (Unknown)
Puce Angreni (Unknown)



Article Info

Publish Date
18 Jan 2026

Abstract

Regression analysis is an analysis to determine the relationship and influence of independent variables on the dependent variable. If the data has a spatial relationship, this analysis has the potential to produce a less accurate model because the regression analysis ignores the influence of the location. One of the data indicated to have a spatial relationship is the open unemployment rate. One spatial analysis that can be used to accommodate spatial relationships is the Geographically Weighted Regression (GWR) model. In the GWR model, a spatial weighting matrix is required whose size depends on the proximity between locations. In this study, two spatial weighting matrix were used: Adaptive Gaussian Kernel and Adaptive Bisquare Kernel. Based on the results of the analysis, it is known that the factors influencing the open unemployment rate in the Riau Islands in 2024 at several locations are the human development index, Economic Growth, and Minimum Wages by Regency/City. Based on the R2 value and AIC value, the best spatial weight matrix produced is the Adaptive Bisquare Kernel weighting function with an R2 value of 93.32% and an AIC value of 15.2835.

Copyrights © 2026






Journal Info

Abbrev

leibniz

Publisher

Subject

Mathematics

Description

Ruang lingkup artikel ilmiah yang dapat diterbitkan dalam Jurnal Leibniz ini adalah sebagai berikut: Geometri dan Aplikasinya, Teori Graf dan Aplikasinya, Riset Operasi dan Aplikasinya, Sistem Dinamik dan Aplikasinya, Model Matematika dan Aplikasinya, Teori Kontrol dan Aplikasinya, Aljabar dan ...