Movatterモバイル変換

saekernel

Propose an area-level, non-parametric regression estimator based onNadaraya-Watson kernel on small area mean. Adopt a two-stage estimationapproach proposed by Prasad and Rao (1990). MSE estimators are notreadily available, so resampling method that called bootstrap isapplied. This package are based on the model proposed in Two stagenon-parametric approach for small area estimation by PushpalMukhopadhyay and Tapabrata Maiti.

Installation

You can install the released version of saekernel fromCRAN with:

install.packages("saekernel")

Authors

Wicak Surya Hasani, Azka Ubaidillah

Maintainer

Wicak Surya Hasani221710052@stis.ac.id

Functions

• saekernel() Produces Small Area EstimationNon-Parametric Based Nadaraya-Watson Kernel
• mse_saekernel() Produces Small Area EstimationNon-Parametric based Nadaraya-Watson Kernel and Bootstrap Mean SquaredError

References

• Mukhopadhyay P, Maiti T. (2004). Two Stage Non-Parametric Approachfor Small Areas Estimation. Proceedings of ASA Section on SurveyResearch Methods: 4058-4065.
• Prasad, N.G.N., and Rao, J.N.K. (1990). “The estimation of the meansquared error of the small area estimators.” Journal of the Americanstatistical association. 85. 163-171.
• Hardle, W. (2002). “Applied non-parametric regression,” CambridgeUniversity Press.
• Indahwati, Sadik K, Nurmasari R. (2008). Pendekatan Metode PemulusanKernel Pada Pendugaan Area Kecil. Makalah Semnas Matematika. UniversitasNegeri Yogyakarta. Yogyakarta.
• Darsyah, M. Y. (2013). Small Area Estimation terhadap PengeluaranPer Kapita di Kabupaten Sumenep Dengan Pendekatan Nonparametrik. JurnalStatistika Volume 1 Nomor 2. Universitas Muhammadiyah Semarang.