Open AccessSpatial model for predicting sugarcane crop productivity using support vector regression
Author(s) -
Abd Gaffar,
Imas Sukaesih Sitanggang
Publication year - 2019
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/335/1/012009
Subject(s) - productivity , production (economics) , agricultural engineering , crop , java , agroforestry , geography , environmental science , forestry , computer science , engineering , economics , macroeconomics , programming language
Sugarcane which is the main ingredient of sugar production has a very high production demand so it must have a balance with the level of productivity. Sugarcane productivity in Indonesia can only reach 5.8 tons/ha which is still relatively low. In recent years sugarcane farmers have felt the decline in sugarcane production due to climate change which has affected the success of sugarcane planting so that sugar production has not been in line with national needs. This study aims to build a prediction model for the productivity of sugarcane crops in each area of Java Island based on climatic factors. The climate data used are collected from Indonesian Agency for Meteorology, Climatology, and Geophysics from 2006 to 2015. The method used is Support Vector Regression (SVR) using a radial base kernel function with a generalized neighbourhood matrix as a spatial model. The prediction model has RMSE of 0.1203954 and correlation of 0.9459743. The results of visualization of sugarcane productivity mapping show that areas in East Java have a higher value of productivity increase than areas in other provinces. Based on the results obtained, the model used has a quite good performance in predicting the productivity of sugarcane.