machine unlearning

{
  "id": "MLC-T-CYB-002563",
  "term": "machine unlearning",
  "field": "Cybersecurity",
  "definition": "A technique that involves selectively removing the influences of specific training data points from a trained machine learning model, such as to remove unwanted capabilities or knowledge in a foundation model, or to enable a user to request the removal of their records from a model. Efficient approximate unlearning techniques may not require retraining the ML model from scratch.",
  "references": [
    "NIST AI 100-2e2025",
    "NIST CSRC Glossary"
  ],
  "url": "https://mlchart.com/terminology/cybersecurity/machine-unlearning/"
}