Pirates are evil? The Marines are righteous? These terms have always changed throughout the course of history! Kids who have never seen peace and kids who have never seen war have different values! Those who stand at the top determine what's wrong and what's right! This very place is neutral ground! Justice will prevail, you say? But of course it will! Whoever wins this war becomes justice! - Donquixote Doflamingo • Pirates are evil? The Marines are righteous? These terms have always changed throughout the course of history! Kids who have never seen peace and kids who have never seen war have different values! Those who stand at the top determine what's wrong

  1. Initial Model Training:

    • Start with a small labeled dataset and train an initial model.
  2. Query Strategy:

    • Use the model to predict on unlabeled data.

    • Identify the most “informative” or “uncertain” data points (e.g., those the model is least confident about).

  3. Labeling:

    • Send these selected data points to a human annotator for labeling.
  4. Model Update:

    • Add the newly labeled data to the training set and retrain the model.
  5. Iteration:

    • Repeat the process until the model achieves satisfactory performance or the labeling budget is exhausted.