PERBANDINGAN METODE BACKPROPAGATION DENGAN MONTE CARLO DALAM MEMPREDIKSI JUMLAH PENDERITA DEMAM BERDARAH DENGUE DI KOTA PEKANBARU

Abstract

Pekanbaru city is one of the cities that have high dengue fever cases. in 2012 recorded 202 cases, 2013 426 cases, 2014 recorded 113 cases and in 2015 recorded as many as 272 cases. The number of each year is always changing, due to various factors that are erratic and other weather changes, it proves that Departmenet of Health of Pekanbaru City has not been able to control the rate of increase in dengue fever patients. The Backpropagation method is a systematic forecasting method for multilayer training, has a strong mathematical basis and is trained by guided methods. With 3 layers, the input layer, the hidden layer and the output layer. Monte Carlo simulation is a probabilistic type of simulation where to obtain the problem solving used sampling method from random process. Monte Carlo experiments on probabilistic elements through random sampling with the aid of generating random numbers.By comparing the two can be known which method is better in predicting the number of patients with dengue fever. The results of data testing with both methods found that Backpropagation method has a low standard deviation compared to Monte Carlo method, so the method of backpropagation is said to be better in predicting the number of dengue hemorrhagic fever patients in Pekanbaru City