JBT Journal
Advanced

Raqamli Transformatsiya va Sun’iy Intellekt

BITKOIN TRANZAKSIYALARIDA ANOMALIYALARNI NAZORATSIZ MASHINAVIY O‘QITISH MODELLARI ASOSIDA ANIQLASH VA ULARNI QO‘SHIMCHA TAHLIL UCHUN SARALASH

Avazov Yusuf Shodiyevich1, Rashidov G'olibjon Xayrullayevich1

1 Islom Karimov nomidagi Toshkent davlat texnika universiteti

Zenodo DOI: 10.5281/zenodo.23656095

Keywords

anomaliyalarni aniqlash tranzaksiya xususiyatlari ECOD blokcheyn nazoratsiz mashinaviy o‘qitish Deep SVDD Bitkoin Isolation forest COPOD

Abstract

Bitkoin blokcheynidagi tranzaksiyalar ochiq saqlansa-da, ularning katta qismi bo‘yicha operatsiya maqsadi yoki xavf darajasini faqat raqamli belgilardan aniqlash qiyin. Ayniqsa, oldindan tasdiqlangan anomaliya belgilari bo‘lmagan ma’lumotlarda qaysi tranzaksiyalarni birinchi navbatda tekshirish lozimligi asosiy muammo hisoblanadi. Tadqiqotda ushbu muammoni nazoratsiz mashinaviy o‘qitish usullari yordamida hal etish imkoniyatlari tahlil qilingan. 2023 yilning yanvaridan 2026 yilning sentabrigacha bo‘lgan davrga tegishli 49 981 ta Bitkoin tranzaksiyasi o‘rganildi. Chiqish qiymati, komissiya, kirishlar va chiqishlar soni modellar uchun asosiy xususiyatlar sifatida tanlanib, qiymatlar orasidagi keskin farqni yumshatish maqsadida logarifmik o‘zgartirildi. Ma’lumotlar vaqt tartibida o‘qitish, validatsiya va test to‘plamlariga ajratildi. Isolation Forest, ECOD, COPOD hamda Deep SVDD g‘oyasiga asoslangan soddalashtirilgan neyron tarmoq prototipi qo‘llanildi. Modellarning baholari alohida chegaralar orqali binar belgilarga o‘tkazildi, natijalar o‘rtasidagi o‘xshashlik Jaccard koeffitsienti bilan o‘lchandi. Validatsiya to‘plamida belgilangan 99,5-protsentil chegaralari test ma’lumotlariga o‘zgartirilmasdan tatbiq etildi.
Test to‘plamida 20 ta tranzaksiya kamida uch model tomonidan, shulardan sakkiztasi esa barcha to‘rt model tomonidan anomal deb belgilandi. Bu guruhda tranzaksiya qiymati, komissiya va kirishlar sonining medianalari boshqa yozuvlarga nisbatan yuqori bo‘ldi. Ayni paytda, modellar orasidagi kelishuv sezilarli farq qildi; bu natijalarning parametrlar va algoritm tanloviga bog‘liqligini ko‘rsatadi. Eng yuqori kelishuv ECOD va COPOD o‘rtasida kuzatildi (Jaccard 0,4500), Isolation Forest va Deep SVDD prototipi o‘rtasida esa u 0,0255 bo‘ldi. Tasdiqlangan anomaliya belgilari yo‘qligi tufayli modellarning aniqligi baholanmadi.
Taklif etilgan yondashuv tahlilchi e’tiborini nisbatan g‘ayrioddiy operatsiyalarga qaratishga yordam beradi, biroq anomaliyani firibgarlik yoki noqonuniy faoliyat bilan tenglashtirmaydi. Bunday xulosa uchun tranzaksiyalarning graf aloqalari, iqtisodiy konteksti va mustaqil tekshiruv natijalari ham zarur

Views: 109 Published: 2026-10-24 Pages: 296-306 Volume: 4 Issue: 5

References

  1. Li, Z., Zhao, Y., Botta, N., Ionescu, C., & Hu, X. (2020). COPOD: Copula-Based Outlier Detection. In Proceedings of the 2020 IEEE International Conference on Data Mining (ICDM) (pp. 1118–1123). IEEE. https://doi.org/10.1109/ ICDM50108.2020.00135
  2. Xiao, B., & Yin, W. (2026). Bidirectional Fusion Heterogeneous Graph Networks for Semi-Supervised Bitcoin Transaction Anomaly Detection in Dynamic Transaction Graphs. PLOS ONE, 21(6), e0351051. https://doi.org/10.1371/journal.pone.0351051
  3. Google Cloud. (n.d.). Bitcoin Transactions Table Schema – BigQuery Public Datasets. https://github.com/GoogleCloudPlatform/knowledge-catalog/blob/main/okf/ bundles/crypto_bitcoin/tables/transactions.md
  4. Ruff, L., Vandermeulen, R., Görnitz, N., Deecke, L., Siddiqui, S. A., Binder, A., Müller, E., & Kloft, M. (2018). Deep One-Class Classification. In Proceedings of the 35th International Conference on Machine Learning (Vol. 80, pp. 4393–4402). PMLR. https://proceedings.mlr.press/v80/ruff18a.html
  5. Ron, D., & Shamir, A. (2013). Quantitative Analysis of the Full Bitcoin Transaction Graph. Financial Cryptography and Data Security, 6–24. Springer. https://doi.org/10.1007/978-3-642-39884-1_2
  6. Liu, F. T., Ting, K. M., & Zhou, Z.-H. (2008). Isolation Forest. In Proceedings of the Eighth IEEE International Conference on Data Mining (pp. 413–422). IEEE. https://doi.org/10.1109/ICDM.2008.17
  7. Huang, W., Li, Y., Xu, Z., Yao, X., & Wan, R. (2025). Improved Deep Support Vector Data Description Model Using Feature Patching for Industrial Anomaly Detection. Sensors, 25(1), 67. https://doi.org/10.3390/s25010067
  8. Hodge, V. J., & Austin, J. (2004). A Survey of Outlier Detection Methodologies. Artificial Intelligence Review, 22(2), 85–126. https://doi.org/10.1023/B:AIRE.0000045502.10941.a9
  9. Akcora, C. G., Li, Y., Gel, Y. R., & Kantarcioglu, M. (2020). BitcoinHeist: Topological Data Analysis for Ransomware Prediction on the Bitcoin Blockchain. Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence (IJCAI), 4439–4445. https://doi.org/10.24963/ijcai.2020/612
  10. Bartoletti, M., Pes, B., & Serusi, S. (2018). Data Mining for Detecting Bitcoin Ponzi Schemes. 2018 Crypto Valley Conference on Blockchain Technology (CVCBT), 75–84. IEEE. https://doi.org/10.1109/CVCBT.2018.00014
  11. Meiklejohn, S., Pomarole, M., Jordan, G., Levchenko, K., McCoy, D., Voelker, G. M., & Savage, S. (2013). A Fistful of Bitcoins: Characterizing Payments Among Men with No Names. Proceedings of the 2013 Internet Measurement Conference, 127–139. ACM. https://doi.org/10.1145/2504730.2504747
  12. Li, Z., Zhao, Y., Hu, X., Botta, N., Ionescu, C., & Chen, G. H. (2023). ECOD: Unsupervised Outlier Detection Using Empirical Cumulative Distribution Functions. IEEE Transactions on Knowledge and Data Engineering, 35(12), 12181–12193. https://doi.org/10.1109/TKDE.2022.3159580
  13. Goldstein, M., & Uchida, S. (2016). A Comparative Evaluation of Unsupervised Anomaly Detection Algorithms for Multivariate Data. PLOS ONE, 11(4), e0152173. https://doi.org/10.1371/journal.pone.0152173
  14. Chandola, V., Banerjee, A., & Kumar, V. (2009). Anomaly Detection: A Survey. ACM Computing Surveys, 41(3), Article 15, 1–58. https://doi.org/10.1145/1541880.1541882