We adapt Snorkel for streaming data. The implementation of drift adaptation strategies are under src/snorkel_incremental.py and src/snorkel_meta.py. The Flink operator of Snorkel is under src/flink_op.py. To reproduce the results we get, please see notebooks directory. The notebooks/snorkel_streaming.ipynb file has the test on credit card fraud task and the notebooks/concept_drift_adaption_snorkel_streaming.ipynb contains the tests for different drift adaptation strategies on the electricity dataset.
-
Notifications
You must be signed in to change notification settings - Fork 1
A weakly supervised learning system for real-time unlabeled streaming data
License
ti1uan/snorkel-streaming
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
A weakly supervised learning system for real-time unlabeled streaming data
Resources
License
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published