Research on optimization of financial fraud detection model based on logistic regression under federated learning framework

Read full text Summarise with VIRAT
Authors: Tan Jixin, Wang Hong
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Logistic regression, as a widely adopted classification algorithm in the financial risk control field, faces the risks of data privacy leakage and bottlenecks in processing efficiency under the traditional centralized training mode. To address these challenges, this study constructs a federated learning framework based on logistic regression, innovatively proposes an optimized model that integrates SMOTE, SOS, and HT. While maintaining the privacy protection advantages of federated learning, the framework achieves dual improvements in model performance and efficiency. Experimental results show that the proposed predictive model exhibits superior predictive performance in handling imbalanced data: the model's AUC value reaches 0.9971, which is 7% higher than that of the traditional federatedlogistic model; the recall rate increases from 32.25% to 97.57%. Additionally, the model reduces the total communication overhead, realizing the coordinated optimization of recognition capability and communication efficiency.

Fields of Study & Subjects

Privacy-Preserving Technologies in Data Imbalanced Data Classification Techniques Financial Distress and Bankruptcy Prediction

Publication details

Journal / source
DOAJ (DOAJ: Directory of Open Access Journals)
Publication date
1 January 2026
Language
ZH