Normalization strategy

{
  "id": "MLC-T-CYB-002884",
  "term": "Normalization strategy",
  "field": "Cybersecurity",
  "definition": "The similarity function can follow one of two normalization strategies, depending on whether the algorithm describes resemblance or containment. For resemblance queries, the number of matching features will be weighed against the total number of features in both objects. In the case of containment queries, the algorithm may disregard unmatched features in the larger of the objects’ two-feature sets.",
  "references": [
    "NIST SP 800-168",
    "NIST CSRC Glossary"
  ],
  "url": "https://mlchart.com/terminology/cybersecurity/normalization-strategy/"
}