Dewi, SintaRatna (2014) Pemodelan Metode Geographically Weighted Ordinaly Logistic Regression Dengan Fungsi Pembobot Adaptive Gaussian Kernel, Adaptive Bisquare Kernel, Dan Adaptive Tricube Kernel. Sarjana thesis, Universitas Brawijaya.
Abstract
Metode yang digunakan untuk menganalisis hubungan antar peubah di mana peubah respon mempunyai skala ordinal dengan peubah prediktor yang bersifat kategori atau kontinu serta memperhatikan faktor geografis tiap pengamatan merupakan Geographically Weighted Ordinal Logistic Regression (GWOLR). Pada penelitian ini dilakukan pemodelan metode Geographically Weighted ordinal logistic Regression (GWOLR) dengan pembobot fungsi adaptive Gaussian kernel, adaptive bisquare kernel, dan adaptive tricube kernel serta menentukan model terbaik pada model Geographically Weighted Ordinal Logistic Regression (GWOLR) antara pembobot fungsi adaptive Gaussian kernel, adaptive bisquare kernel, dan adaptive tricube kernel pada data Indeks Pembangunan Manusia (IPM) wilayah Jawa Timur tahun 2010. Berdasarkan nilai Normalized Root Mean Square Error (NRMSE) disimpulkan bahwa model GWOLR dengan pembobot adaptive bisquare kernel adalah model yang sesuai digunakan untuk memodelkan data IPM Jawa Timur tahun 2010 dibandingkan dengan metode GWOLR dengan fungsi pembobot adaptive gaussian kernel dan pembobot adaptive tricube kernel.
English Abstract
The method used to analyze the relationship between variables in which the response variable has an ordinal scale with the predictor variables are continuous and considering the category or geographical factors of each observation is Geographically Weighted Ordinal Logistic Regression (GWOLR). In this research, a method of modeling ordinal logistic Geographically Weighted Regression (GWOLR) with adaptive Gaussian kernel weighting function, adaptive bisquare kernel, and adaptive tricube kernel and determine the best model in the model Geographically Weighted Ordinal Logistic Regression (GWOLR) between adaptive Gaussian kernel weighting function, bisquare adaptive kernel and adaptive kernel data tricube Human Development Index (HDI) in East Java in 2010 Based on the Normalized Root Mean Square Error (NRMSE) concluded that GWOLR models with adaptive weighted bisquare kernel is an appropriate model is used to model the data HDI East Java in 2010 compared with GWOLR method with adaptive gaussian kernel weighting function and adaptive weighting tricube kernel.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/MIPA/2014/339/051405837 |
| Subjects: | 500 Natural sciences and mathematics > 510 Mathematics |
| Divisions: | Fakultas Matematika dan Ilmu Pengetahuan Alam > Matematika |
| Depositing User: | Budi Wahyono Wahyono |
| Date Deposited: | 15 Sep 2014 13:43 |
| Last Modified: | 21 Oct 2021 05:42 |
| URI: | http://repository.ub.ac.id/id/eprint/153973 |
Preview |
Text
PEMODELAN_GEOGRAPHICALLY_WEIGHTED_ORDINALY_LOGISTIC_REGRESSI.pdf Download (4MB) | Preview |
Actions (login required)
 |
View Item |