Alatoo Academic Studies, 2025, т. 20, № 2, с. 536-542
FORECASTING CURRENCY EXCHANGE RATES IN KYRGYZSTAN USING THE ARIMA MODEL: DEVELOPMENT OF AN AUTOMATED SYSTEM IN JAVA
Аннотация · Abstract
Transparent, data-driven prediction systems are essential, as evidenced by the growing need for precise currency exchange rate forecasts in developing nations. The design and implementation of a Java-based forecasting system that uses the Autoregressive Integrated Moving Average (ARIMA) model to forecast changes in the Kyrgyz som's (KGS) exchange rate in relation to key foreign currencies is shown in this article. This project includes a fully manual realisation of the ARIMA model in Java, incorporating all three components — AR, I, and MA — together with automatic parameter selection, in contrast to standard implementations in Python or R. The National Bank of the Kyrgyz Republic (NBKR) gives historical data to the system, which then processes it for stationarity and uses a RESTful API to provide short- and medium-term projections. High predictive accuracy for stable time periods is demonstrated by evaluation utilising MAE and RMSE. In addition to providing a useful answer for Kyrgyz analysts, companies, and policymakers, this work lays the groundwork for future hybrid models and more comprehensive economic forecasts.
Сведения · Details
- DOI 10.17015/aas.2025.252.45
- Журнал · Journal Alatoo Academic Studies
- Язык · Language английский
- Цитирования (Crossref) · Citations 0