| dc.contributor |
Fakultas Matematika dan Ilmu Pengetahuan Alam |
|
| dc.creator |
Junianingsih, Safni Chusnaifah |
|
| dc.creator |
Karyana, Yayat |
|
| dc.date |
2020-08-25 |
|
| dc.date.accessioned |
2021-03-15T03:45:36Z |
|
| dc.date.available |
2021-03-15T03:45:36Z |
|
| dc.identifier |
http://karyailmiah.unisba.ac.id/index.php/statistika/article/view/23581 |
|
| dc.identifier |
10.29313/.v6i2.23581 |
|
| dc.identifier.uri |
http://hdl.handle.net/123456789/28890 |
|
| dc.description |
Abstract. Regression analysis is one of the analytical tools used to determine the effect of multiple predictor variables (X) on response variables (Y). The approach in regression analysis is divided into two, parametric approaches and nonparametric approaches. On nonparametric regression analysis, the shape of the regression curve is unknown, the data arega expected to look for its own estimation form so that it has high flexibility. Estimation of regression functions is performed with the Nadaraya Watson kernel estimator using Gaussian kernel functions. In this method requires bandwidth (h) or finer parameters as a balance controller between the smoothness of the function and the suitability of the function of the data. Optimum bandwidth (h) is obtained by minimizing the Generalized Cross Validation (GCV) value. Based on the analysis, obtained in a simple linear regression model obtained a Mean Square Error (MSE) value of 552976772 and a Standard Error (SE) of 24437,98. While in the kernel nonparametric regression model, the optimum bandwidth (h) is 0,50, Mean Square Error (MSE) is 96832714, and the Standard Error (SE) value is 10226,4. So it can be concluded that the kernel nonparametric regression model is better than a simple linear regression model.Keywords: Generalized Cross Validation (GCV), Baby Birth, Nadaraya Watson, Kernel RegressionAbstrak. Analisis regresi merupakan salah satu alat analisis yang digunakan untuk mengetahui pengaruh dari beberapa variabel prediktor (X) terhadap variabel respon (Y). Pendekatan dalam analisis regresi dibagi menjadi dua, yaitu pendekatan parametrik dan pendekatan nonparametrik. Pada analisis regresi nonparametrik bentuk kurva regresi tidak diketahui, data diharapkan mencari sendiri bentuk estimasinya sehingga memiliki fleksibilitas yang tinggi. Estimasi fungsi regresi dilakukan dengan estimator kernel Nadaraya Watson menggunakan fungsi kernel Gaussian. Metode ini membutuhkan bandwidth (h) atau parameter penghalus sebagai pengontrol keseimbangan antara kemulusan fungsi dan kesesuaian fungsi terhadap data. Bandwidth (h) optimum diperoleh dengan meminimumkan nilai Generalized Cross Validation (GCV). Berdasarkan analisis diperoleh pada model regresi linear sederhana diperoleh nilai Mean Square Error (MSE) sebesar 552976772 dan niai Standard Error (SE) sebesar 24437,98. Sedangkan pada model regresi nonparametrik kernel diperoleh bandwidth (h) optimum sebesar 0,50, Mean Square Error (MSE) sebesar 96832714, dan nilai Standard Error (SE) sebesar 10226,4. Sehingga dapat disimpulkan bahwa model regresi nonparametrik kernel lebih baik daripada model regresi linear sederhana.Kata Kunci: Generalized Cross Validation (GCV), Kelahiran Bayi, Nadaraya Watson, Regresi Kernel. |
|
| dc.format |
application/pdf |
|
| dc.language |
eng |
|
| dc.publisher |
Universitas islam Bandung |
|
| dc.relation |
http://karyailmiah.unisba.ac.id/index.php/statistika/article/view/23581/pdf |
|
| dc.rights |
Copyright (c) 2020 Prosiding Statistika |
|
| dc.source |
Prosiding Statistika; Vol 6, No 2, Prosiding Statistika (Agustus, 2020); 312-319 |
|
| dc.source |
Prosiding Statistika; Vol 6, No 2, Prosiding Statistika (Agustus, 2020); 312-319 |
|
| dc.source |
2460-6456 |
|
| dc.source |
10.29313/.v6i2 |
|
| dc.subject |
Statistika |
|
| dc.subject |
Generalized Cross Validation (GCV), Baby Birth, Nadaraya Watson, Kernel Regression |
|
| dc.title |
Regresi Nonparametrik Kernel dalam Pemodelan Jumlah Kelahiran Bayi di Jawa Barat Tahun 2017 |
|
| dc.type |
info:eu-repo/semantics/article |
|
| dc.type |
info:eu-repo/semantics/publishedVersion |
|
| dc.type |
Peer-reviewed Article |
|
| dc.type |
kuantitatif |
|