PERBANDINGAN PERAMALAN MENGGUNAKAN DOUBLE EXPONENTIAL SMOOTHING HOLT DAN FUZZY TIME SERIES MARKOV CHAIN UNTUK DATA INFLASI INDONESIA

Ica Rahma Kurniawati (Department of Statistics, Faculty of Science and Mathematics , Diponegoro University)
Suparti Suparti (Departemen Statistika, Fakultas Sains dan Matematika, Universitas Diponegoro)
Tarno Tarno (Departemen Statistika, Fakultas Sains dan Matematika, Universitas Diponegoro)

Jurnal Gaussian
Vol 15, No 1 (2026): Jurnal Gaussian

Inflation is one measure of a country’s economic condition. The inflation rate in Indonesia had decreased to 1.33% in June 2021 when the Covid-19 pandemic was hitting. Forecast of the inflation rate is needed for monitoring economic condition. The Double Exponential Smoothing Holt method can be used on data that contains trend elements and uses two smoothing parameters in the analysis, while the Fuzzy Time method Series Markov Chain is a development of the Fuzzy Time Series method with the Markov chain concept which does not have any conditions in its analysis. The results of forecasting analysis from the Double Exponential Smoothing Holt method obtained an MAE value of 0.35957, while from the Fuzzy Time Series Markov Chain method obtained an MAE value of 0.22462. Thus, the Fuzzy Time Series Markov Chain model is a better model in forecasting than the Double Exponential Smoothing Holt method. That model is reused to analyze forecasting and get a MAPE value of 8.79% which means that the Fuzzy Time Series Markov Chain model predicts Indonesia’s inflation data very well.