In reality, there are very few identifiable virtual currency financial crime problems, this paper is studied from the micro level of the transaction subject and the macro level of the transaction currency, first of all, by building a machine learning algorithm to classify the elliptic data set node, the tree integration model has achieved better prediction effect in predicting abnormal Bitcoin transactions, which is more suitable for using this model to build an abnormal transaction detection system. Then, through the event research method, the impact of the collapse of larger transaction entities on the price fluctuation of currencies is studied, and it is found that some currencies are more affected by the platform failure event, and some currencies are less affected.

Xinhao He, Chenyu Wang, Yu Zou, Wenyu Zhang. Research on Virtual Currency Trading Behavior under Financial Technology Innovation. Academic Journal of Business & Management (2023) Vol. 5, Issue 17: 167-172. https://doi.org/10.25236/AJBM.2023.051725.

[1] Zhen Yuanyun. Legal Risks and Regulation of Virtual Currency [J]. Academic Journal of Humanities & Social Sciences, 2022, 5(17). 

[2] Dimitrios Koutmos. Bitcoin returns and transaction activity [J]. Economics Letters, 2018, 167. 

[3] Chipere, Mike. Virtual currency as an inclusive monetary innovation for the unbanked poor[J]. Electronic Commerce Research & Applications, 2018, 28:37-43.DOI:10.1016/j.elerap.2018.01.004.

[4] Chawla N V, Bowyer K W, Hall L O, et al. SMOTE: synthetic minority over-sampling technique [J]. Journal of artificial intelligence research, 2002, 16(1): 321–357. 

[5] Seth Armitage. Event Study Methods and Evidence On Their Performance [J]. Journal of Economic Surveys, 1995, 9(1). 

[6] Hogarth K, Dawson D. Implementing e-Learning in Organisations: What E-Learning Research Can Learn From Instructional Technology (IT) and Organisational Studies (OS) Innovation Studies[J]. International journal on E-learning, 2008, 7(1):p.87-105.