Scholarlink

<p>This study develops a data-driven model of USD/JPY exchange rate volatility using reservoir computing, a recurrent neural network technique capable of capturing nonlinear deterministic dynamics. Widely applied in the physical sciences and engineering over the past decade, this approach is employed here to investigate financial time series. Results demonstrate that the constructed model can produce reliable forecasts several weeks ahead under certain initial conditions, suggesting the presence of deterministic recurrent behavior in exchange rate volatility.</p>

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

Identifying deterministic behavior in exchange rate volatility time series using machine learning

  • 期刊论文

  • 发表日期：

    2026年5月3日

查看全文
  • Makoto Muto,
  • Tamotsu Onozaki,
  • Yoshitaka Saiki

摘要

This study develops a data-driven model of USD/JPY exchange rate volatility using reservoir computing, a recurrent neural network technique capable of capturing nonlinear deterministic dynamics. Widely applied in the physical sciences and engineering over the past decade, this approach is employed here to investigate financial time series. Results demonstrate that the constructed model can produce reliable forecasts several weeks ahead under certain initial conditions, suggesting the presence of deterministic recurrent behavior in exchange rate volatility.