Anggraeni, Siska Oktaviana Dwi (2022) Small Area Estimation Dengan Thinned Spatial Point Process Untuk Menduga Rata-Rata Pengeluaran Per Kapita Per Bulan Tingkat Desa/Kelurahan Di Kabupaten Kendal Jawa Tengah. Masters thesis, Institut Teknologi Sepuluh Nopember Surabaya.
![[thumbnail of 6003201023-Master_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
6003201023-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 November 2024. Download (2MB) | Request a copy |
Abstract
Pengeluaran per kapita per bulan digunakan sebagai dasar perhitungan kemiskinan sekaligus indikator untuk mengukur tingkat ketimpangan pendapatan penduduk dengan cara membandingkannya antar wilayah. Namun demikian, data pengeluaran per kapita per bulan belum tersedia hingga ke level desa/kelurahan. Metode small area estimation (SAE) menjadi salah satu solusi karena ukuran sampel Survei Sosial Ekonomi Nasional (SUSENAS) yang kurang representatif. Adanya informasi geo-referensi pada unit sampel SUSENAS membuka kesempatan baru untuk memodelkan rata-rata pengeluaran per kapita per bulan dengan thinned spatial point process untuk dapat dibandingkan dengan model spasial SAE berbasis area. Kelebihan model thinned spatial point process dalam SAE adalah: (1) mempertimbangkan dependensi dan distribusi spasial pengeluaran per kapita per bulan, (2) dapat memodelkan heterogeneity melalui efek variabel prediktor, dan (3) mampu melakukan estimasi rata-rata pengeluaran per kapita per bulan pada level yang sangat detail (koordinat) secara kontinyu. Penelitian ini berfokus pada pengembangan metode SAE dengan thinned spatial point process menggunakan model marked LGCP pada rata-rata pengeluaran per kapita per bulan tingkat desa/kelurahan di Kabupaten Kendal. Berdasarkan hasil penelitian, SAE dengan pendekatan thinned spatial point process melalui model marked LGCP memberikan nilai prediksi yang terbaik dengan nilai RMSE terkecil dibandingkan SAE konvensional dengan pendekatan area. Lebih lanjut, hasil prediksi model marked LGCP mampu menangkap variasi spasial rata-rata pengeluaran per kapita per bulan di setiap desa/kelurahan yang tidak mampu ditangkap oleh model SEBLUP serta dapat disajikannya nilai hasil prediksi pada setiap unit rumah tangga melalui informasi koordinat
======================================================================================================================================
Expenditure per capita per month is used as a basis for poverty measurement as well as an indicator to measure the level of population income inequality by its value between regions. However, the expenditure per capita data is not yet available at the district level due to the lack of representation of the sample size of the National Socio-Economic Survey (SUSENAS). The small area estimation (SAE) method is one of the solutions. The existence of geo-reference information in the SUSENAS sample unit embark new opportunity to predict the average expenditure per capita at the district level using a thinned spatial point process to be compared with the area-based spatial SAE model. The advantages of the thinned spatial point process model in SAE are: (1) considering the dependencies and spatial distribution of the average expenditure per capita per month of the household, (2) capturing heterogeneity through the effects of predictor variables, and (3) being able to estimate the average expenditure per capita per month at a very detailed level (coordinates) continuously. This study focuses on developing a SAE method with a thinned spatial point process using marked LGCP model on the average household expenditure per capita per month at the district level in Kendal Regency. The results showed that SAE with the thinned spatial point process approach through the marked LGCP model provided the best predicted value with the minimum RMSE compared to conventional SAE with spatial approach. Furthermore, the predicted results of the marked LGCP model can capture the spatial variation of the average expenditure per capita per month in each district that cannot be captured by the SEBLUP model. The marked LGCP model can present the predicted value for each household unit through coordinate information
| Item Type: | Thesis (Masters) |
|---|---|
| Additional Information: | RTSt 519.544 Ang s-1 2022 |
| Uncontrolled Keywords: | pengeluaran per kapita per bulan, thinned spatial point process, SAE, SEBLUP, marked LGCP; expenditure per capita per month, thinned spatial point process, SAE, SEBLUP, marked LGCP |
| Subjects: | H Social Sciences > HA Statistics > HA31.7 Estimation |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49101-(S2) Master Thesis |
| Depositing User: | EKO BUDI RAHARJO |
| Date Deposited: | 28 Mar 2023 03:12 |
| Last Modified: | 28 Mar 2023 03:12 |
| URI: | http://repository.its.ac.id/id/eprint/97816 |
Actions (login required)
 |
View Item |