<![CDATA[Before sugar cane was milled by the factory, the first process is analysis of sugar cane maturity. The best sugar cane condition to be ground is mature cane that can be seen from several factors such as garden area, age, stem diameter, the average segment per stem and the average length per stem. These factors are used as attributes in the research conducted. To simplify the process, then we proposed this research on the prediction of sugar cane harvest time. With so much data being used and repeated processes, it will be difficult to process manually and takes a long time. In addition, the manual process does not close the possibility of an increasing error. This research uses a combination of genetic algorithm and backpropagation in the process of predicting the harvest time. Genetic algorithms are the best solution used to optimize prediction results by weight selection and bias. Backpropagation method is used to calculate Mean Square Error (MSE) value, which will be used in calculation of fitness value and also on prediction of data test. In this research will be done five kinds of testing, as follows generation test, population size test, test combination of crossover rate and mutation rate, testing of learning rate and testing of Average Forecasting Error Rate (AFER). The result of this research are predictions of harvest time, the value of fitness and AFER. The best result is result of AFER value is 0,0205%.]]>
Copyrights © 2017
