<![CDATA[The train is a kind of massive land transport with a lot of users, base on the results presented by Statistics for Safety Index and service reached 4.09 from 5 in year 2014, also supported by the fact that exposed by the daily Tempo (www.bisnis.tempo.co) indicating that the train users from time to time inCreased. However, with the inCrease in the number of passengers on top of the train without any prediction will be bad for the train in Indonesia. For this need a method of predicting the results that can be answerable, using popular methods such as artificial neural network Backpropagation and optimizations to do in determining the initial weights (W) with Using numbered variables 800 for the population, 20 for a number of generations, the composition of the value of Mr = 0.3 and Cr = 0.7, with the main variant of the Backpropagation artificial neural network that consists of multiple iterations is 100 and a value of Alpha is 0.9, also with dataset on a monthly basis, start from January 2006 to June 2017 in timeseries form data, with 100 training data pattern as initial data and 10 pattern of test data of last data. So the result is the level of precision based on error value (MSE) results 0.065869861 from the results of the hybridization method backpropagation artificially neural networks using a genetic algorithm, while without using the hybridization error value is 0.072517977.]]>
Copyrights © 2019
