Active Learning

Active Learning is a supervised learning paradigm that iteratively trains predictors and selects the most informative training samples to optimize the model, achieving better model performance and higher prediction accuracy with less data, thereby improving the efficiency of configuration discovery and the model's generalization ability.