[new feature] add (another model-free) data valuation method: complexity-gap score #972
Labels
enhancement
New feature or request
help-wanted
We need your help to add this, but it may be more challenging than a "good first issue"
Cleanlab's data valuation module: https://docs.cleanlab.ai/master/cleanlab/datalab/internal/issue_manager/data_valuation.html
Currently solely supports Data Shapely based on K-nearest-neighbor models.
Goal: add more efficient Data Valuation methods as options for this module.
One good one to consider (should benchmark against Cleanlab's current KNN Data Shapely) is the complexity-gap method published in this paper:
DATA VALUATION WITHOUT TRAINING OF A MODEL
https://arxiv.org/abs/2301.00930
The text was updated successfully, but these errors were encountered: