data heterogeneity

The variation and distribution differences across datasets, which can affect model performance and generalization, particularly in distributed learning scenarios where data may be non-IID (independent and identically distributed).

23 papers