Scholarlink

<p>This research study evaluates the effectiveness of the NeuroFuzzy and Copula models for modelling and predicting the exchange rates of the cryptocurrencies namely, Bitcoin (BTC), Ethereum (ETH), and Litecoin (LTC) using environmental (CO<sub>2</sub>) and economical (Petroleum Price, PP) factors. Comparative analysis reveals that the Copula model consistently outperforms the NeuroFuzzy model by achieving significantly lower error metrics. The Copula exhibits superior accuracy, especially for BTC and LTC where it achieves near-zero MSE, highlighting its robustness in modelling the complex dependencies. Conversely, the NeuroFuzzy, while effective for capturing non-linearity, exhibits higher error metrics, suggesting limitations in handling intricate interdependencies, especially for ETH. These findings underscore the efficacy of Copula models in enhancing forecasting accuracy by capturing tail and dependency structures. Future research should explore the dynamic Copula framework and integrate additional market sentiment indicators and blockchain-specific variables to further refine predictive models for the high-volatility cryptocurrency market.</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

Integrating NeuroFuzzy and Copula Model for Robust Cryptocurrency Exchange Rate Forecasting Using Environmental and Economic Data

  • 期刊论文

  • 发表日期：

    2025年7月11日

查看全文
  • Imran Ali Khan,
  • Sami Ur Rahman

摘要

This research study evaluates the effectiveness of the NeuroFuzzy and Copula models for modelling and predicting the exchange rates of the cryptocurrencies namely, Bitcoin (BTC), Ethereum (ETH), and Litecoin (LTC) using environmental (CO2) and economical (Petroleum Price, PP) factors. Comparative analysis reveals that the Copula model consistently outperforms the NeuroFuzzy model by achieving significantly lower error metrics. The Copula exhibits superior accuracy, especially for BTC and LTC where it achieves near-zero MSE, highlighting its robustness in modelling the complex dependencies. Conversely, the NeuroFuzzy, while effective for capturing non-linearity, exhibits higher error metrics, suggesting limitations in handling intricate interdependencies, especially for ETH. These findings underscore the efficacy of Copula models in enhancing forecasting accuracy by capturing tail and dependency structures. Future research should explore the dynamic Copula framework and integrate additional market sentiment indicators and blockchain-specific variables to further refine predictive models for the high-volatility cryptocurrency market.