Zahiroh, Nazmatuz (2017) Regresi Robust S-Estimation dan MM-Estimation untuk Pemodelan Luas Panen Padi dengan Indikator Iklim di Sentra Produksi Padi Jawa Timur. Undergraduate thesis, Institut Tekologi Sepuluh Nopember.
Preview |
Text
1315105037-Undergraduate_Theses.pdf - Published Version Download (3MB) | Preview |
Abstract
Padi merupakan tanaman pokok nasional yang dikonsumsi oleh sebagian masyarakat Indonesia. Menurut indikator pertanian 2016, Provinsi Jawa Timur memberikan sumbangan terbesar bagi produksi tanaman pangan nasional. Padi merupakan tanaman yang paling rentan terhadap perubahan pola curah hujan, yang akan berimbas pada luas area tanam dan panen, produktivitas, dan kualitas hasil panen. Identifikasi hubungan antara luas panen padi dan curah hujan dibutuhkan di sentra produksi padi sebagai antisipasi kegagalan panen akibat perubahan iklim. Penelitian ini ingin mendapatkan model luas panen padi yang terbaik di beberapa kabupaten di Jawa Timur menggunakan regresi robust S-estimation dan MM-estimation. Hasil identifikasi hubungan antara curah hujan dan luas panen padi menunjukkan pola hubungan linier positif. Pemodelan dengan robust S-estimation dan MM-estimation menghasilkan lebih banyak variabel signifikan dan persentase ketepatan prediksi yang lebih tinggi daripada regresi OLS. Model luas panen padi Kabupaten Jember dan Bojonegoro lebih optimal menggunakan MM-estimation, sedangkan model luas panen padi di Kabupaten Banyuwangi, Ngawi, dan Lamongan lebih optimal menggunakan S-estimation. Koefisien determinasi yang dihasilkan pada setiap periode yaitu 1%-50% untuk periode 1, 0%-40% untuk periode 2, dan 24%-87% untuk periode 3.
========================================================================
Rice is a national staple plant for Indonesian. According to agricultural indicators 2016, East Java Province gives the largest contribution of national food crop production. Rice is the most vulnerable plant with fluctuate change in rainfall patterns, it gives impacts on cropping and harvesting area, productivity, and harvest yield quality. Identification of the relationship between rice harvest area and rainfall is needed, especially in rice production centers, to anticipate crop failure due to climate changes. This study aims to get the best model of rice harvest area in some districts of East Java by using robust regression with S-estimation and MM-estimation. The result of relationship identification between rainfall and rice harvest area shows a positive linear relationship pattern. Modelling with robust S-estimation and MM-estimation resulted in more significant variables and higher predicted accuracy percentages than OLS regression. The model of rice harvest area in Jember and Bojonegoro districs is more optimal using MM-estimation, while the model of rice harvest area in Banyuwangi, Ngawi, and Lamongan districts is more optimal using S-estimation. Determination coefficient that resulted at each period are 1%-50% for period 1, 0%-40% for period 2, and 24%-87% for period 3.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Uncontrolled Keywords: | Curah Hujan, Luas Panen Padi, Robust S-Estimation, Robust MM-Estimation |
| Subjects: | H Social Sciences > HA Statistics Q Science > Q Science (General) |
| Divisions: | Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Nazmatuz Zahiroh . |
| Date Deposited: | 08 Jan 2018 03:51 |
| Last Modified: | 05 Mar 2019 07:51 |
| URI: | http://repository.its.ac.id/id/eprint/47831 |
Actions (login required)
 |
View Item |