Jurnal Penelitian Pertanian Tanaman Pangan
Vol 33, No 1 (2014): April 2014

Pemodelan Pengukuran Luas Panen Padi Nasional Menggunakan Generalized Autoregressive Conditional Heteroscedastic Model (GARCH)

Iqbal, Teuku Achmad (Mahasiswa Pascasarjana Departemen Statistika IPB Jl. Meranti, Kampus IPB Dramaga, Bogor)
Sadik, Kusman (Dosen Departemen Statistika IPB)
Sumertajaya, I Made (Dosen Departemen Statistika IPB)



Article Info

Publish Date
30 Apr 2014

Abstract

This study was aimed to build a model for the estimation of national harvested area of rice by incorporating element of variant heterogeneity and the influence of asymmetry factors on time series data using five types of GARCH models, namely: symmetric GARCH, exponential asymmetric GARCH, quadratic asymmetric GARCH, Threshold GARCH, and non-linear asymmetric GARCH. Those models were compared and evaluated, and then the best model was used to predict the accuracy of the national rice harvested area. The results showed that two types of GARCH had significant coefficient, indicating the validity of the model. Those models were symmetric GARCH and quadratic GARCH models. Based on the value of mean absolute percentage error (MAPE) for the twelve month periods ahead, quadratic GARCH model was better than the symmetric GARCH model. Furthermore, based on the value of mean absolute deviation (MAD) and mean square error (MSE), quadratic GARCH model also seemed to be a better model than symmetric GARCH model. The best model can be used to predict the harvested area in the subsequent year.

Copyrights © 2014






Journal Info

Abbrev

jpptp

Publisher

Subject

Agriculture, Biological Sciences & Forestry

Description

Jurnal Penelitian Pertanian Tanaman Pangan merupakan publikasi yang memuat makalah ilmiah primer hasil penelitian tanaman pangan (padi dan palawija). Redaksi menerima makalah dari peneliti Puslitbang Tanaman Pangan, Balai Pengkajian Teknologi Pertanian (BPTP), maupun perguruan tinggi. Makalah yang ...