PREDIKSI TINGKAT INFLASI DI INDONESIA BERDASARKAN PERUBAHAN HARGA BAHAN BAKAR MINYAK (BBM) DENGAN MENGGUNAKAN ALGORITMA GENETIKA; PREDICTION OF INFLATION RATE IN INDONESIA BASED ON THE CHANGE OF FUEL OIL PRICE BY USING GENETIC ALGORITHM

Sari, Arum Novia, Fizah

2015 | Skripsi | FMIPA UGM

Government policy to raise the price of fuel oil (BBM), especially premium in recent months led to a variety of effects. Impact of a very influential one of which is the rate of inflation. If the fuel price increases, the inflation rate will also increase. Therefore, the government should be able to predict the rate of inflation based on the increase in fuel prices. Of the prediction results obtained, the government will be able to prepare fiscal policies and non-fiscal to face all the risks that will happen and maintain economic stability in Indonesia. Genetic algorithm is one of the great computational methods to predict. By using a genetic algorithm that applies natural generation technique that is evolution, a solution can be sought by promising computational model. The process of genetic algorithms in this study starts from the generation of the initial population, the selection of the parent with the roulette wheel selection, crossover process with whole arithmetic crossover, random mutation process up to the selection of survivors in the form of steady-state updates are done well to get a prediction solution. Testing the system in this study can be predicted inflation rate for the past and the future. In testing the predictions of inflation system found other factors besides the price premium. Another factor is worth 3.83. With the reduction of other factors, the test results predicted in the reign of Susilo Bambang Yudhoyono during a 10-year gained increased accuracy from 42.98% to 62.16%. Accuracy testing for prediction of the month January 2015 by 87%. While testing accuracy for the prediction of the month February 2015 amounted to 97%.

Kata Kunci : prediction; inflation; fuel prices; genetic algorithms