SMALL AREA ESTIMATION TERHADAP PENGELUARAN PER KAPITA PADA LEVEL KECAMATAN DI KABUPATEN SUMENEP DENGAN PENDEKATAN KERNEL–BOOTSTRAP

DARSYAH, MOH YAMIN (2013) SMALL AREA ESTIMATION TERHADAP PENGELUARAN PER KAPITA PADA LEVEL KECAMATAN DI KABUPATEN SUMENEP DENGAN PENDEKATAN KERNEL–BOOTSTRAP. Masters thesis, Institut Teknologi Sepuluh Nopember.

[img]
Preview
Text
1311201010-Abstract.pdf - Published Version

Download (180kB) | Preview
[img] Text
1311201010-Master_Theses.pdf - Published Version
Restricted to Repository staff only

Download (940kB) | Request a copy
[img]
Preview
Text
1311201010-Conclution.pdf - Published Version

Download (170kB) | Preview

Abstract

Pendugaan area kecil dengan teknik pendugaan tak langsung memerlukan asumsi adanya hubungan linier antara rataan area kecil dengan variabel penyerta. Jika tidak ada hubungan linier antara rataan area kecil dan variabel penyerta maka tidak tepat ‘meminjam kekuatan’ dari area lain dengan menggunakan model linier dalam pendugaan tak langsung. Untuk mengatasi hal tersebut dikembangkan pendekatan nonparametrik. Salah satu pendekatan nonparametrik yang digunakan adalah pendekatan Kernel-Bootstrap. Pendugaan tak langsung dengan pendekatan SAE Kernel-Bootstrap digunakan untuk menduga pengeluaran per kapita pada level kecamatan di Kabupaten Sumenep . Evaluasi hasil pendugaan dilakukan dengan membandingkan nilai RRMSE (Relative Root Mean Square Error) penduga langsung dengan nilai RRMSE (Relative Root Mean Square Error) penduga SAE Kernel-Bootstrap. Hasil pendugaan SAE Kernel-Bootstrap lebih presisi dibanding pendugaan langsung ===================================================================================================== Small area estimation with indirect estimation techniques require the assumption of linear relationship between the average small area with concomitant variables. If there is no linear relationship between the average small area and the concomitant variable is not appropriate to 'borrowing strength' from other areas by using a linear model in the indirect estimation. To overcome this nonparametric approach is developed. One approach is used the nonparametric kernel-based approach. Indirect Estimation with SAE Kernel- Booststrap in this study is applied to estimate per capita expenditure in several districs in the sumenep regency. Evaluation of the estimation made by comparing the value RRMSE (Relative Root Mean Square Error) estimator RRMSE directly with the value RRMSE (Relative Root Mean Square Error) estimator RRMSE of SAE Kernel-Bootstrap estimator. The result of SAE Kernel-Bootstrap estimation is more precition than result of directly estimation.

Item Type: Thesis (Masters)
Additional Information: RTSt 519.544 Dar s
Uncontrolled Keywords: SAE, Kernel, Bootstrap, RRMSE
Subjects: Q Science > QA Mathematics
Q Science > QA Mathematics > QA278.2 Regression Analysis
Q Science > QA Mathematics > QA278 Cluster Analysis
Divisions: Faculty of Mathematics and Science > Statistics > (S2) Master Theses
Depositing User: Mr. Tondo Indra Nyata
Date Deposited: 05 Jan 2017 04:29
Last Modified: 05 Jan 2017 04:29
URI: http://repository.its.ac.id/id/eprint/1326

Actions (login required)

View Item View Item