Komparasi Kinerja Fuzzy Time Series dengan Model Rantai Markov dalam Meramalkan PDRB Provinsi Bali

This paper aimed to elaborates and compares the performance of Fuzzy Time Series (FTS) model with Markov Chain (MC) model in forecasting the Gross Regional Domestic Product (GDRP) of Bali Province. Both methods were considered as forecasting methods in soft modeling domain. The data used was quarterly data of Bali’s GDRP for year 1992 through 2013 from Indonesian Bureau of Statistic at Denpasar Office. Inspite of using the original data, rate of change from two consecutive quarters was used to model. From the in-sample forecasting conducted, we got the Average Forecasting Error Rate (AFER) for FTS dan MC models as much as 0,78 percent and 2,74 percent, respectively. Based-on these findings, FTS outperformed MC in in-sample forecasting for GDRP of Bali’s data.