Active Learning

Active learning is a supervised learning paradigm that iteratively trains predictors and selects training samples that can improve model accuracy and optimize configurations, achieving better performance with less data. This method has significant application value in situations where data labeling costs are high and resources are limited.