Scholarlink

In accordance with Bank Indonesia Regulation Number 12/16/PBI/2010, Article 1, Bank Indonesia, as an independent state institution, must establish a system capable of predicting foreign exchange rates in real-time. A technology-based approach is used to improve accuracy in analyzing and forecasting exchange rate movements activities by currency type, as well as to formulate policy recommendations. To achieve this, the study employs machine learning-based data analysis methods, focusing on seven major foreign currencies: JPY, KRW, SGD, USD, HKD, and CNY. The selection of monetary indicators BI-Rate, Consumer Price Index (CPI) inflation, inflation targets, foreign exchange reserves, JISDOR, and IndONIA is based on correlation coefficient calculations. Algorithms applied include SARIMAX and Multiverse LSTM. Prediction errors are evaluated using metrics such as Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE), which serve as references for formulating policy recommendations to improve Bank Indonesia’s regulations.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prediction of Exchange Rates of Foreign Currency Using Machine Learning as a Policy Formulation Tool SISMONTAVAR

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Firman Chandra Alamsyah,
  • Helni Mutiarsih Jumhur,
  • Arief Arianto Hidayat

摘要

In accordance with Bank Indonesia Regulation Number 12/16/PBI/2010, Article 1, Bank Indonesia, as an independent state institution, must establish a system capable of predicting foreign exchange rates in real-time. A technology-based approach is used to improve accuracy in analyzing and forecasting exchange rate movements activities by currency type, as well as to formulate policy recommendations. To achieve this, the study employs machine learning-based data analysis methods, focusing on seven major foreign currencies: JPY, KRW, SGD, USD, HKD, and CNY. The selection of monetary indicators BI-Rate, Consumer Price Index (CPI) inflation, inflation targets, foreign exchange reserves, JISDOR, and IndONIA is based on correlation coefficient calculations. Algorithms applied include SARIMAX and Multiverse LSTM. Prediction errors are evaluated using metrics such as Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE), which serve as references for formulating policy recommendations to improve Bank Indonesia’s regulations.