One-class support vector machines for detecting population drift in deployed machine learning medical diagnostics
Abstract Machine learning (ML) models are increasingly being applied to diagnose and predict disease, but face technical challenges such as population drift, where the training and real-world deployed data distributions differ. This phenomenon can degrade model performance, risking incorrect diagnos...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-04-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-94427-x |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Be the first to leave a comment!