Implementasi Fuzzy Time Series Markov Chain untuk Peramalan Inflasi

Abstract

The Fuzzy Time Series Markov Chain method has the ability to manage complex and unstable Time Series data, and provide more accurate predictions by minimising forecasting errors. In this study, the Fuzzy Time Series Markov Chain method is implemented in forecasting inflation in North Sumatra. The results show that the FTS-MC method produces a Mean Squared Error (MSE) of 0.0097 and a Mean Absolute Percentage Error (MAPE) value of 0.02%, which indicates an accuracy rate of 99.98%. In comparison, the Fuzzy Time Series method without Markov Chain produces an MSE of 0.37 and a MAPE of 0.040%, with an accuracy rate of 99.96% and the use of ARIMA produces an MSE of 1.325 and a MAPE of 0.892%. Based on these results, the FTS-MC method is proven to be superior in predicting inflation with very small deviations. Inflation forecasting for the next seven periods shows fluctuations, indicating dynamic changes in inflation in North Sumatra. This method can be an effective tool in inflation forecasting to support better economic decision making.