Throwing Away Data Improves Worst-Class Error in Imbalanced Classification.
Martín ArjovskyKamalika ChaudhuriDavid Lopez-PazPublished in: CoRR (2022)
Keyphrases
- imbalanced datasets
- decision trees
- data collection
- data sets
- database
- training data
- raw data
- data sources
- data quality
- multi class
- image data
- knowledge discovery
- data analysis
- data structure
- target class
- multiple classes
- class discrimination
- multiclass classification
- data distribution
- training samples
- feature set
- feature extraction