Budiani, Jauhara Rana (2015) Analisis Hubungan Dan Pemodelan Luas Panen Padi Dengan Indikator Enso Di Kabupaten Bojonegoro Melalui Pendekatan Copula Dan Regresi Robust M-Estimation. Undergraduate thesis, Institut Technology Sepuluh Nopember.
Preview |
Text
1313105020-Undergraduate Thesis.pdf - Published Version Download (1MB) | Preview |
Abstract
Bojonegoro merupakan salah satu daerah lumbung padi di
Indonesia. Perairan sawah padi di Bojonegoro masih bergantung pada
ketersediaan air hujan. Jika curah hujan tinggi akan berakibat
kebanjiran, sedangkan pada musim kemarau akan mengalami
kekeringan. Upaya untuk meminimalkan kerugian akibat kebanjiran
atau kekeringan adalah dengan mengetahui karakteristik hubungan
serta pemodelan produksi padi dengan faktor iklim yaitu El-Nino
Southern Oscillation (ENSO). Data luas panen padi dan indikator
ENSO memiliki nilai yang ekstrim sehingga asumsi kenormalan
seringkali terlanggar. Penelitian ini menggunkan pendekatan Copula
untuk mengidentifikasi hubungan antar variabel, karena tidak terlalu
ketat dalam asumsi distribusi variabel. Pemodelan yang digunakan
untuk mengetasi data ekstrim atau outlier adalah regresi robust Mestimation.
Hasil penelitian menunjukkan bahwa dependensi antara luas
panen padi dan indikator ENSO 3.4 sebagian besar mengikuti Copula
Clayton yang memiliki tail dependensi di bawah, artinya kejadian
ekstrim SST Nino 3.4 bernilai rendah dan luas panen padi di
Bojonegoro yang diperoleh akan semakin kecil maka hubungannya
semakin erat, terutama pada periode 3. Pada pemodelan dengan regresi
Robust M-Estimation menunjukkan nilai koefisien determinasi yang
masih kecil yaitu pada periode 1 sebesar 33,94%, periode 2 sebesar
16,63%, dan periode 3 sebesar 16,72%.
======================================================================================================
Bojonegoro is one of the largest rice production in Indonesia. Aquatic rice
fields in Bojonegoro is still dependent on the availability of rainwater. If
heavy rainfall will cause flooding, while in the dry season will suffer
drought. Efforts to minimize losses due to flood or drought is to determine
the characteristics of the relationship and rice production modeling with
climate factors that El-Nino Southern Oscillation (ENSO). Rice harvested
area of data and indicators of ENSO has an extreme value that normality
assumption is often violated. This study using the Copula approach to
identify relationships between variables, because it is not too tight in the
variable distribution assumptions. Modeling is used to treat extreme data or
outlier is robust regression M-estimation. The results showed that the
relationship between the rice harvested area and ENSO indicators Nino 3.4
region largely followed the Clayton Copula tail that has a dependency on
the bottom, meaning that extreme events Nino 3.4 SST low value and the
rice harvested area in Bojonegoro obtained will be smaller then do more
closely, especially on subground 3. On Robust regression modeling with MEstimation
shows the coefficient of determination that a child is at
oubground 1 is 32.68%, subground 2 is 19.82%, and subground 3
adalah14,18%.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSSt 519.536 Bud a |
| Uncontrolled Keywords: | ENSO, Copula, Robust M-Estimation, luas panen padi |
| Subjects: | Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression |
| Divisions: | Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Mr. Tondo Indra Nyata |
| Date Deposited: | 06 Dec 2018 02:05 |
| Last Modified: | 06 Dec 2018 02:05 |
| URI: | http://repository.its.ac.id/id/eprint/60078 |
Actions (login required)
 |
View Item |