| Abstrakt: |
The main objective of this study is the forecasting of USD exchange rate using the Fuzzy Time Series. The data used is from January 2001 to March 2022. The methods used are Time-Invariant Fuzzy Time Series (FTS), Chen FTS and Cheng FTS, using first-order and second-order. FTS is a forecasting technique that makes use of fuzzy sets and rules-based logic. Then, based on the Mean Absolute Percentage Error (MAPE) number, the degree of prediction accuracy is determined. In order to determine which approach is best for this case study, the MAPE values of all five strategies are then compared. The results showed that Time-Invariant FST produced 7,39%, First-Order Chen FTS produced MAPE of 5,13% and First-Order Cheng FTS produced 5,46%. Second-Order Chen FTS produced 2,18% and Second-Order Cheng FTS produced 2,09 %. The results of this study indicate that Second-Order Cheng FTS produces good accuracy and can be used to predict the USD exchange rate. [ABSTRACT FROM AUTHOR] |
