COMPARATIVE ANALYSIS OF FUZZY TIME SERIES MARKOV CHAIN AND FUZZY TIME SERIES CHENG MODELS IN INFLATION PREDICTION KUDUS REGENCY

Abstract

Inflation is one of the important economic indicators that reflects price stability and people's purchasing power. Unpredictable inflation fluctuations require accurate forecasting methods to support planning and policy-making, especially at the regional level. This study aims to compare the performance of Markov Chain Fuzzy Time Series (FTS) and Cheng FTS in predicting inflation in Kudus Regency and to determine the most effective and efficient model. The data used is monthly inflation data for Kudus Regency, which is analyzed through the stages of determining the universe of discourse, interval formation, fuzzification, fuzzy logic relationship formation, and defuzzification. The accuracy level of the model is evaluated using Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). The results showed that the Markov Chain FTS model performed better than the Cheng FTS model. The Markov Chain FTS produced an MAE value of 0.1811 and an RMSE of 0.2371, which were smaller than those of the Cheng FTS, which produced an MAE of 0.2659 and an RMSE of 0.3656. This advantage is due to the Markov Chain FTS's ability to utilize transition adjustments between states, making it more adaptive to data dynamics. Thus, it can be concluded that the Markov Chain FTS is the most effective and efficient model for predicting inflation in Kudus Regency.

Keywords: Fluctuations, Fuzzy Time Series, Markov chain, MAE, RMSE.