Sabrina, Nezalfa (2026) PENERAPAN METODE MULTISCALE GEOGRAPHICALLY WEIGHTED REGRESSION UNTUK PEMODELAN TINGKAT PARTISIPASI ANGKATAN KERJA PEREMPUAN DI PULAU JAWA TAHUN 2024. Undergraduate thesis, UPN Veteran Jawa Timur.
|
Text (Cover)
22083010067-cover.pdf Download (807kB) | Preview |
|
|
Text (BAB 1)
22083010067-bab1.pdf Download (97kB) | Preview |
|
![[img]](https://repository.upnjatim.ac.id/style/images/fileicons/text.png) |
Text (BAB 2)
22083010067-bab2.pdf Restricted to Repository staff only until 8 July 2029. Download (387kB) |
|
|
Text (BAB 3)
22083010067-bab3.pdf Restricted to Repository staff only until 8 July 2029. Download (545kB) |
|
|
Text (BAB 4)
22083010067-bab4.pdf Restricted to Repository staff only until 8 July 2029. Download (1MB) |
|
|
Text (BAB 5)
22083010067-bab5.pdf Download (15kB) | Preview |
|
|
Text (Daftar pustaka)
22083010067-daftarpustaka.pdf Download (96kB) | Preview |
|
|
Text (Lampiran)
22083010067-lampiran.pdf Restricted to Repository staff only until 8 July 2029. Download (99kB) |
Abstract
The Female Labor Force Participation Rate (FLFPR) in Indonesia remains the third lowest in ASEAN at 54.8%, with stark disparities across regions in Java Island. This study aims to apply the Multiscale Geographically Weighted Regression (MGWR) method to model female LFPR across 119 regencies/cities in Java Island in 2024, compare the performance of four kernel functions (Adaptive Gaussian, Adaptive Bisquare, Adaptive Exponential, and Adaptive Triangular), and develop a Streamlit-based User Interface to visualize the modeling results. MGWR was selected for its ability to capture spatially varying influence scales by assigning different bandwidths to each independent variable, and to address multicollinearity among independent variables. The independent variables used include the Regency/City Minimum Wage (MSEs), the number of poor people, the Gender Inequality Index (GII), Gross Regional Domestic Product (GRDP), the percentage of women in parliament, the number of women managing households, and the Gender Development Index (GDI). Data were obtained from official 2024 publications of Statistics Indonesia (BPS). Assumption testing results indicate significant spatial heterogeneity and spatial autocorrelation, reinforcing the justification for the MGWR approach over global OLS regression, which performed poorly. Based on evaluation using R², AICc, MSE, and RMSE criteria, the MGWR model with the Adaptive Bisquare kernel was selected as the best model with R² = 0.7210, outperforming both the GWR model and MGWR models with the other three kernels. The developed User Interface facilitates the analysis workflow with kernel selection features, local equation tables per region, and interpretation of analysis results.
| Item Type: | Thesis (Undergraduate) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Contributors: |
|
||||||||||||
| Subjects: | Q Science > QA Mathematics | ||||||||||||
| Divisions: | Faculty of Computer Science > Departemen of Data Science | ||||||||||||
| Depositing User: | Ms NEZALFA SABRINA | ||||||||||||
| Date Deposited: | 08 Jul 2026 03:45 | ||||||||||||
| Last Modified: | 08 Jul 2026 03:45 | ||||||||||||
| URI: | https://repository.upnjatim.ac.id/id/eprint/54797 |
Actions (login required)
 |
View Item |