EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Numerical Signal Design for Crypto Intelligence

Sai Zhang, Chongbin Luo, Annika Cai, Yeran Lu

Journal of Global Economic Insights · 2025

DOI: 10.71204/b2c2rb55 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Global Economic Insights, 2025. Verfasst von Sai Zhang und 3 weiteren Autor:innen. Die Arbeit nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study presents a novel framework that bridges deterministic numerical algorithms with computational finance to support interpretable machine learning applications in cryptocurrency analysis. By applying Newton’s method, the Trapezoidal Rule, and a hybrid Euler–Adams-Bashforth solver, we generate structured numerical features that capture convergence, integration precision, and differential dynamics, respectively. These feature vectors are constructed from both synthetically generated sequences and real AVAX-USD log return data, enabling a direct comparison between theoretical numerical behavior and market-driven fluctuations. A deterministic labeling rule, based on the parity of the integer sum of the features, defines class boundaries with geometric regularity, allowing the K-Nearest Neighbors classifier to operate in a feature space shaped by mathematically grounded transformations. The results reveal that classical numerical methods, when applied to financial time series, produce stable, class-separable patterns that are well-suited for local classification and boundary interpretation. Through decision boundary visualization, error convergence analysis, and joint feature interaction plots, the paper demonstrates that numerical approximation techniques can illuminate latent structural signals in crypto price behavior. This framework advances the integration of scientific computing and data-driven finance, offering a new paradigm for understanding digital asset dynamics through the lens of deterministic modeling.

Zitieren

APA

Zhang, S., Luo, C., Cai, A., & Lu, Y. (2025). Numerical Signal Design for Crypto Intelligence. Journal of Global Economic Insights. https://doi.org/10.71204/b2c2rb55

Vancouver

Zhang S, Luo C, Cai A, Lu Y. Numerical Signal Design for Crypto Intelligence. Journal of Global Economic Insights. 2025. doi:10.71204/b2c2rb55

BibTeX
@article{Zhang2025,
  author = {Zhang, Sai and Luo, Chongbin and Cai, Annika and Lu, Yeran},
  title = {Numerical Signal Design for Crypto Intelligence},
  journal = {Journal of Global Economic Insights},
  year = {2025},
  doi = {10.71204/b2c2rb55}
}
Kostenlos starten →