AcceptanceThreshold

AcceptanceThreshold

is an option that specifies the minimum threshold at which a result is considered acceptable.

Details

  • Typical settings include:
  • Automaticautomatic threshold
    vala numerical value typically between 0 and 1
  • Typically, if the confidence for a particular result is below the setting for AcceptanceThreshold, it will be rejected.
  • If all results are rejected, typically Missing[] is returned.
  • In Classify, Predict and anomaly detection functions, the value of AcceptanceThreshold is used to define anomalies and corresponds to a RarerProbability threshold.

Examples

Basic Examples  (4)

Specify the acceptance threshold while detecting faces:

Lower the threshold:

ImageInstanceQ uses AcceptanceThreshold->0.5:

With a higher recognition threshold, objects may not be identified:

ImageIdentify uses an automatic acceptance threshold:

Specify a custom threshold:

If no identification is above the threshold, a Missing object is returned:

Delete anomalies with the default AcceptanceThreshold value:

Delete anomalies with a higher threshold:

Introduced in 2019
 (12.0)