Regresi Nonparametrik Spline Truncated untuk Memodelkan Persentase Unmet Need di Kabupaten Gresik

Sholicha, Camelia Nanda (2018) Regresi Nonparametrik Spline Truncated untuk Memodelkan Persentase Unmet Need di Kabupaten Gresik. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[img]
Preview
Text
06211645000038-Undergraduate_Theses.pdf - Accepted Version

Download (4MB) | Preview

Abstract

Kabupaten Gresik merupakan wilayah yang mengalami penurunan sehingga persentase unmet need pada tahun 2017 menjadi urutan terendah kedua di Provinsi Jawa Timur yaitu sebesar 15,6%. Persentase tersebut masih tinggi dibandingkan target pemerintah sebesar 7,03% di tiap Kabupaten/Kota. Selanjutnya, pemerintah Kabupaten Gresik diharapkan dapat meningkatkan pencapaian Kabupaten Gresik dalam program KB dan fertilitas dengan mempertimbangkan faktor-faktor yang mempengaruhi seperti petugas penyuluh informasi, tempat pelayanan KB, persentase keluarga miskin dan persentase pendidikan terakhir lebih dari SMA yang ditamatkan kepala keluarga. Sehingga metode yang sesuai digunakan dalam penelitian yaitu analisis regresi nonparametrik Spline Truncated karena data tidak membentuk sebuah pola tertentu sehingga menggunakan titik knot optimum berdasarkan nilai Generalized Cross Validation (GCV) minimum. Dalam penelitian ini, karakteristik kecamatan dengan persentase unmet need di Kabupaten Gresik sangat tinggi dan tidak memenuhi SPM. Model regresi nonparametrik spline truncated terbaik adalah dengan menggunakan kombinasi titik knot yaitu satu titik knot pada variabel cakupan PKB/PLKB, satu titik knot pada variabel cakupan tempat pelayanan KB, dua titik knot pada variabel persentase keluarga miskin, serta dua titik knot pada variabel persentase pendidikan terakhir lebih dari SMA yang ditamatkan oleh KK. Nilai koefisien determinasi yang dihasilkan dari model ini adalah sebesar 88,09%. ================================================================================================ Gresik Regency is an area that has decreased the percentage of unmet need so that in 2017 it becomes the second lowest in East Java province which is 15.6%. The percentage is still high compared to the government target of 7.03% in each regency/city. Furthermore, the government of Gresik Regency is expected to increase the achievement of Gresik Regency in KB program and fertility by considering influencing factors such as information extension officer, place of family planning service, percentage of poor family and the last education percentage more than SHS. So the appropriate method used is nonparametric Spline Truncated regression analysis because the data does not form a certain pattern so using the optimum knot point based on the minimum Generalized Cross Validation (GCV). In this research, the characteristics of sub-districts with unmet need percentage in Gresik Regency are very high and do not meet minimum service standards. The best spline truncated nonparametric regression model is to use a combination of knot points ie one knot point on the coverage variable of the field officer, one knot point on the coverage variable of service family planning, two knots on the percentage of poor families, as well as two points of knots on the last percentage of education variables over high school that were rescued by the head of the family. The coefficient of determination resulted from this model is 88,09%.

Item Type: Thesis (Undergraduate)
Uncontrolled Keywords: GCV, Regresi Nonparametrik, Spline Truncated, Titik Knot, Unmet Need
Subjects: H Social Sciences > HA Statistics > HA31.3 Regression. Correlation
Divisions: Faculty of Mathematics, Computation, and Data Science > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User: Camelia nanda sholicha
Date Deposited: 21 Jul 2021 23:35
Last Modified: 21 Jul 2021 23:35
URI: https://repository.its.ac.id/id/eprint/57741

Actions (login required)

View Item View Item