<![CDATA[The exchange rate of the Indonesian Rupiah against the US Dollar experiences frequent fluctuations, making economic forecasting and financial planning more difficult. This study aims to enhance exchange rate prediction accuracy by combining Fuzzy Time Series with Markov Chain probability transitions. The approach is grounded in the idea that probabilistic modeling of state changes improves the representation of dynamic currency behavior. Using daily IDR/USD data from April 2020 to March 2025, the methodology involves two main steps: fuzzifying historical exchange rate data into linguistic variables, and applying a Markov Chain to compute transition probabilities between these fuzzy states. The model’s forecasting accuracy is evaluated using mean absolute percentage error. Results show that the hybrid model achieves a lower error rate of 0.50%, compared to 0.61% using conventional Fuzzy Time Series alone. This demonstrates the hybrid model’s ability to capture both sudden market changes and stable patterns effectively. The findings suggest that the integration of Markov Chain transitions significantly improves the predictive performance of fuzzy-based models. In conclusion, this hybrid method provides a practical and reliable forecasting tool for financial analysts and policymakers. Future research could include additional economic indicators and explore alternative probability weighting methods to further enhance model accuracy.]]>
Copyrights © 2025
