PEMODELAN PRODUKSI PADI DI INDONESIA DENGAN PENDEKATAN MIXED GEOGRAPHICALLY WEIGHTED REGRESSION
No Thumbnail Available
Date
2021-03
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This paper discusses about rice production modeling in Indonesia with Mixed
Geographically Weighted Regression (MGWR) approach. The MGWR model is a
combination of multiple linear regression models and Geographically Weighted
Regression (GWR), which model variables that are global and local. The estimated
parameters of the MGWR model are obtained using the Weighted Least Square (WLS)
method. The optimum bandwidth is selected by using the Cross Validation (CV) method.
Rice production in Indonesia is influenced by factors such as harvest area, rainfall,
fertilizer, and farmers. MGWR model produces variables of local nature is the area of
harvest, while the variables that are global are rainfall, fertilizer and farmers. MGWR
model is a best model used to model rice production data in Indonesia based on Akaike
Information Criteria Corrected value
Description
Keywords
Rice production, Mixed Geographically Weighted Regression, Weighted Least Square, Cross Validation, Akaike Information Criterion Corrected