org.jpmml.evaluator

Interface HasProbability

All Superinterfaces:

CategoricalResultFeature, ResultFeature

All Known Implementing Classes:

NeuronClassificationMap, NodeClassificationMap, ProbabilityClassificationMap, VoteClassificationMap

public interface HasProbability
extends CategoricalResultFeature

A marker interface for probabilistic classification results.

Probability represents a degree of certainty that the sample belongs to a particular category. Probabilities are required to sum to 1 across all categories. Ideally, the probability of the predicted category should approach 1.0, and the probabilities of all other categories should approach 0.0.

Classification is a supervised learning task. The set of all categories is defined by the valid values of the target field. The set of known categories may be smaller than the set of all categories. It is assumed that the probability of missing categories is 0.0.

See Also:

ResultFeatureType.PROBABILITY

Method Summary

Methods

Modifier and Type	Method and Description
Double	getProbability(String value) Gets the probability of the specified category.

Methods inherited from interface org.jpmml.evaluator.CategoricalResultFeature

getCategoryValues

Method Detail

getProbability

Double getProbability(String value)

Gets the probability of the specified category.

Returns:

A probability in the range from 0.0 to 1.0. The probability of an unknown category is 0.0.

See Also:

CategoricalResultFeature.getCategoryValues()