WP3: Human-in-the-loop approach
Model training
Initial model training and generalization of patterns from data.
Curriculum Learning
Organizing datasets by difficulty, expert disagreements measure complexity.
Active Learning
System queries experts for the most informative cases to label.
Interactive Machine Learning
Experts provide fluid feedback, correcting errors and offering unprompted insights.
Enhanced model performance
Achieving a more robust, generalizable, and efficient AI model.
This WP integrates advanced Human-in-the-Loop (HITL) approaches, including Curriculum Learning and Active Learning, evolving towards Interactive Machine Learning (IML). The core objective is to significantly enhance model performance through continuous, expert-driven feedback.