Peramalan Indeks Harga Konsumen Kota Medan Menggunakan Metode Fuzzy Time Series Markov Chain

Abstract

The Fuzzy Time Series Markov Chain (FTSMC) method is used to forecast time series data that contains uncertainty, based on fuzzy set theory and the concept of Markov chains. This forecasting process involves four main stages: fuzzification of historical data, formation of fuzzy groups and the relationships between intervals, construction of a Markov transition matrix from the established fuzzy relationships, and finally, forecasting future values using state transition probabilities. The FTSMC model is applied to estimate the Consumer Price Index (CPI) of Medan City for the period from April 2020 to March 2025. Method accuracy is evaluated using three metrics: Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE), which aim to assess the performance of the projection results. The evaluation shows that the FTSMC approach is capable of producing highly accurate predictions with MAE of 0.613; RMSE of 1.526; and MAPE of 0.576%. These values indicate that the Fuzzy Time Series Markov Chain method used in this study yields relatively low forecasting errors for the CPI data of Medan City. Therefore, this method can be considered a viable alternative for forecasting economic time series that are dynamic and involve elements of uncertainty.