Pemodelan Kawasan Panen Kelapa Sawit Malaysia Menggunakan Model Logistik

Abstract

Palm oil and related products is Malaysia's second largest export. Palm oil state expenditures in 2010 are projected to reach approximately 64,282,738 tons within a year. However, these results still have not reached the stage of maximum when seen from an oil palm plantation areas in Peninsular Malaysia, Sabah and Sarawak covering an area of 4,691,160 hectares. Of the area, approximately 2,807 million hectares of oil palm plantations planted by the private sector (59.8%) while 20.2% is managed by the Government (FELDA 675.167 ha, 80.262 Risda and Felcra 160.832), 321.947 ha (7.1%) by the local government and the rest covering an area of 540.194 ha (12.9%) are currently working on. This study discusses the entire area of oil palm planted area compared to the harvest, the number of workers factors, yields and prices. The model used in this research is the method of logistics, with the probit model, logit model and complementary log-log function (complementary log-log) to 39 locations palm observations drawn from Jabatan Perangkaan Malaysia. Models are selected based on statistical test that involves the value of AIC (Akaike information criterion) of proc logistic method, the value of the method Genmod Proc devians as well as the value of p, the reference is taken to determine the best model in this study. After comparison of the statistical test is selected, indicating that the model Complement function logs is a better model than the models probit and logit models.
Keywords: AIC, devians, genmod proc, logistic proc, palm oil