edu.stanford.nlp.neural.rnn

Class RNNCoreAnnotations.PredictedClass

java.lang.Object

edu.stanford.nlp.neural.rnn.RNNCoreAnnotations.PredictedClass

All Implemented Interfaces:

CoreAnnotation<java.lang.Integer>, TypesafeMap.Key<java.lang.Integer>

Enclosing class:

RNNCoreAnnotations

public static class RNNCoreAnnotations.PredictedClass
extends java.lang.Object
implements CoreAnnotation<java.lang.Integer>

Get the argmax of the class predictions. The predicted classes can be an arbitrary set of non-negative integer classes, but in our current sentiment models, the values used are on a 5-point scale of 0 = very negative, 1 = negative, 2 = neutral, 3 = positive, and 4 = very positive.

Constructor Summary

Constructors

Constructor and Description
PredictedClass()

Method Summary

Modifier and Type	Method and Description
java.lang.Class<java.lang.Integer>	getType() Returns the type associated with this annotation.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

PredictedClass

public PredictedClass()

Method Detail

getType

public java.lang.Class<java.lang.Integer> getType()

Description copied from interface: CoreAnnotation

Returns the type associated with this annotation. This method must return the same class type as its value type parameter. It feels like one should be able to get away without this method, but because Java erases the generic type signature, that info disappears at runtime.

Specified by:

getType in interface CoreAnnotation<java.lang.Integer>