Open AccessOPTIMASI PARAMETER MODel AUTOREGRESSIVE MENGGUNAKAN ALGORITMA PARTICLE SWARM OPTIMIZATION
Author(s) -
Setyoko Prismanu Ramadhan,
Hasbi Yasin,
Suparti Suparti
Publication year - 2019
Publication title -
jurnal gaussian : jurnal statistika undip
Language(s) - English
Resource type - Journals
ISSN - 2339-2541
DOI - 10.14710/j.gauss.v8i2.26666
Subject(s) - autoregressive integrated moving average , particle swarm optimization , autoregressive model , series (stratigraphy) , box–jenkins , time series , mean absolute percentage error , mean squared error , nonlinear autoregressive exogenous model , algorithm , mathematical optimization , mathematics , computer science , statistics , paleontology , biology
Box-Jenkins ARIMA method is a linear model in time series analysis which is widely used in various fields. One estimation method for Box-Jenkins ARIMA model is OLS method which aims to minimize the number of squared errors. This method is not effective when applied to time series data that is random, nonlinear and non-stationary. In this study discussed the alternative method of the PSO algorithm as an parameter optimization of the ARIMA model. PSO algorithm is an optimization method based on the behavior of a flock of birds or fish. The main advantage of the PSO algorithm is having a simple, easy to implement and efficient concept in calculations. This method is applied to data from PT Perusahaan Gas Negara shares. The results of both methods will be compared. In the AR model (1) the value of MSE is 0.532 and MAPE is 0.993. Meanwhile, the PSO algorithm obtained MSE 0.531 and MAPE 0.988. It was found that the PSO algorithm resulted in smaller MSE and MAPE values and could provide better results.Keywords : Time Series Analysis, Autoregressive, PSO