EUVIMEDEUVIMED entdecken
Crossref

Adaptive Ai-Based Algorithm Selection Framework for The Discrete Logarithm Problem

Abdulloh Hamidov Abduqodir o‘g‘li

American Journal of Applied Science and Technology · 01.07.2026 · 6 (7) · 47-53

DOI: 10.37547/ajast/volume06issue07-06 ↗

Kurz gefasst

Fachartikel in der Zeitschrift American Journal of Applied Science and Technology, online erschienen am 01.07.2026. Verfasst von Abdulloh Hamidov Abduqodir o‘g‘li.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study investigates how to automatically choose the fastest algorithm for solving a given instance of the Discrete Logarithm Problem (DLP). The DLP underlies many public-key cryptographic systems, yet no single algorithm is optimal for every instance: Pohlig–Hellman is highly effective when the group order is smooth, Pollard’s Rho and the Kangaroo method are suitable for prime-order and interval-restricted cases, and Index Calculus dominates over large prime fields. In this study, algorithm selection is formulated as a supervised classification task. For each instance, a compact set of number-theoretic features—bit length, largest prime factor, smoothness ratio, interval width, and simplified cost estimates—is computed, and three tree-based models, namely Decision Tree, Random Forest, and XGBoost, are trained on a balanced synthetic dataset of 24,000 instances. XGBoost achieved 94.6% accuracy and a 94.3% macro F1-score. Feature-importance analysis showed that the smoothness ratio and the largest prime factor are the most decisive predictors for algorithm choice. The findings confirm that a lightweight machine learning model can serve as a reliable and interpretable tool both for selecting DLP solvers and for assessing the real hardness of group parameters.

Zitieren

APA

o‘g‘li, A. H. A. (2026). Adaptive Ai-Based Algorithm Selection Framework for The Discrete Logarithm Problem. American Journal of Applied Science and Technology, 6(7), 47-53. https://doi.org/10.37547/ajast/volume06issue07-06

Vancouver

o‘g‘li AHA. Adaptive Ai-Based Algorithm Selection Framework for The Discrete Logarithm Problem. American Journal of Applied Science and Technology. 2026;6(7):47-53. doi:10.37547/ajast/volume06issue07-06

BibTeX
@article{ogli2026,
  author = {o‘g‘li, Abdulloh Hamidov Abduqodir},
  title = {Adaptive Ai-Based Algorithm Selection Framework for The Discrete Logarithm Problem},
  journal = {American Journal of Applied Science and Technology},
  year = {2026},
  volume = {6},
  number = {7},
  pages = {47--53},
  doi = {10.37547/ajast/volume06issue07-06}
}
Kostenlos starten →