Definition
The set of techniques and interventions applied during data collection, model training, or post-processing to reduce unfair and systematic biases in AI systems.
In short: The set of techniques and interventions applied during data collection, model training, or post-processing to reduce unfair and systematic biases in AI systems.