Experiments on real datasets find that balancing methods increase predictive multiplicity in Rashomon sets of models, measured via ambiguity, discrepancy, and a new obscurity metric.
M., Seiffert, C., Van Hulse, J., Napolitano, A., Folleco, A.: Learn- ing with limited minority class data
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
An Experimental Study on the Rashomon Effect of Balancing Methods in Imbalanced Classification
Experiments on real datasets find that balancing methods increase predictive multiplicity in Rashomon sets of models, measured via ambiguity, discrepancy, and a new obscurity metric.