A dynamic receptive field and improved feature fusion approach for federated learning in financial credit risk assessment
Ruiheng Li, Yue Cao, Yuhang Shu, Jia Guo, Binghua Shi, Jiaojiao Yu, Yi Di, Qiankun Zuo, Hao Tian · 2024
Read A dynamic receptive field and improved feature fusion approach for federated learning in financial credit risk assessment by Ruiheng Li, Yue Cao, Yuhang Shu, Jia Guo, Binghua Shi, Jiaojiao Yu, Yi Di, Qiankun Zuo, Hao Tian (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-11532548.txt.
Description
Li and colleagues propose in 2024 a federated learning design for judging credit risk without pooling private financial data. Federated systems train locally and merge models centrally, but communication costs balloon with uneven data like default records. Their method adjusts the model's receptive field during feature extraction. It also builds a fusion scheme that blends local and global features with fewer transmissions.
In experiments, the design kept accuracy while cutting communication rounds by about four-fifths. Dynamic kernels allocated weights to suit varied default patterns. The authors present the framework as a cheaper path to privacy-preserving credit scoring.
Catalog record
| CREATOR | Ruiheng Li, Yue Cao, Yuhang Shu, Jia Guo, Binghua Shi, Jiaojiao Yu, Yi Di, Qiankun Zuo, Hao Tian |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 54 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |