Wulandari, Novita Dwi (2017) Pemodelan Mixed Geographically Temporally Weighted Regression (Mgtwr) Pada Kasus Produksi Tanaman Padi Di Provinsi Jawa Timur. Sarjana thesis, Universitas Brawijaya.
Abstract
Model Mixed Geographically Temporally Weighted Regression (MGTWR) merupakan model yang memuat beberapa peubah prediktor bersifat global (diasumsikan konstan untuk seluruh titik pengamatan) sedangkan yang lainnya bervariasi secara spasial (lokasi) dan secara temporal (waktu). Padi adalah komoditas utama yang berperan sebagai pemenuh kebutuhan pokok karbohidrat bagi penduduk. Tujuan dari penelitian ini, yaitu mengembangkan model MGTWR dengan fungsi pembobot gaussian, bisquare, dan tricube kernel serta menentukan fungsi pembobot yang paling baik pada produksi tanaman padi di Jawa Timur tahun 2011-2015. Berdasarkan hasil analisis model MGTWR yang terbentuk terdapat 38 kabupaten/kota yang memiliki nilai prediksi tentang produksi tanaman padi yang berbeda sesuai dengan lokasi dan waktu. Model MGTWR dengan fungsi pembobot gaussian kernel lebih baik digunakan untuk memodelkan produksi tanaman padi di Jawa Timur tahun 2011-2015 dibandingkan model MGTWR fungsi pembobot bisquare kernel dan tricube kernel. Model MGTWR terbaik yang terbentuk menggambarkan faktor-faktor yang berpengaruh signifikan terhadap produksi tanaman padi di Jawa Timur tahun 2011-2015. Faktor-faktor tersebut bervariasi di setiap tahun dari kabupaten/kota di Jawa Timur, namun secara umum faktor yang berpengaruh terhadap produksi tanaman padi adalah luas panen.
English Abstract
Mixed Geogaphically Temporally Weighted Regression (MGTWR) model was a model contained some predictor variable to be global (assumpted constant for all monitoring point) whereas the others had variation in a spacial manner (location) and in a temporal manner (time). Rice plant was a prominent commodity which had rule to fullfil the main needs of carbohydrate for society. The purpose of this research was to develop MGTWR model with function of gaussian, bisquare, and tricube kernel weighting and to decide the best weighting of rice plant production in East Java on 2011-2015. Based on the result of analysis MGTWR model formed that consist with 38 regency/city which had prediction score about rice plant that were different with the location and the time. MGTWR model with gaussian kernel weighting’s function was better to use for modeling the rice plant production in East Java on 2011-2015 than MGTWR model with bisquare kernel and tricube kernel weighting’s function. The best of MGTWR model was formed describing factors that had significant effect toward rice plant production in East Java on 2011-2015. Those factors had varieties in each years from regency/city in East Java, but in general the factor that affect toward the rice plant production is the wide of the harvest.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/FMIPA/2017/263/051705762 |
| Uncontrolled Keywords: | bisquare, gaussian, MGTWR, padi, tricube |
| Subjects: | 500 Natural sciences and mathematics > 519 Probabilities and applied mathematics > 519.5 Statistical mathematics > 519.53 Descriptive statistics, multivariate analysis, analysis of variance and covariance > 519.536 Regression analysis |
| Divisions: | Fakultas Matematika dan Ilmu Pengetahuan Alam > Statistika |
| Depositing User: | Nur Cholis |
| Date Deposited: | 24 Oct 2017 01:47 |
| Last Modified: | 24 Nov 2021 02:38 |
| URI: | http://repository.ub.ac.id/id/eprint/4235 |
Preview |
Text
Skripsi Full.pdf Download (4MB) | Preview |
Actions (login required)
 |
View Item |