org.deeplearning4j.eval

Class Evaluation

java.lang.Object

org.deeplearning4j.eval.Evaluation

public class Evaluation
extends Object

Evaluation metrics: precision, recall, f1

Author:

Adam Gibson

Constructor Summary

Constructors

Constructor and Description
Evaluation()

Method Summary

Methods

Modifier and Type	Method and Description
void	addToConfusion(int real, int guess) Adds to the confusion matrix
int	classCount(int i) Returns the number of times the given label has actually occurred
void	eval(org.jblas.DoubleMatrix realOutcomes, org.jblas.DoubleMatrix guesses) Collects statistics on the real outcomes vs the guesses.
double	f1()
double	f1(int i) Calculate f1 score for a given class
void	incrementFalseNegatives()
void	incrementFalsePositives(int i)
void	incrementTruePositives()
int	numtimesPredicted(int label) Returns the number of times a given label was predicted
int	numTimesPredicted(int actual, int predicted) Gets the number of times the given class was predicted for the given predicted label
double	precision()
double	precision(int i) Returns the precision for a given label
double	recall() Returns the recall for the outcomes
String	stats()

Methods inherited from class java.lang.Object

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

Constructor Detail

Evaluation

public Evaluation()

Method Detail

eval

public void eval(org.jblas.DoubleMatrix realOutcomes,
        org.jblas.DoubleMatrix guesses)

Collects statistics on the real outcomes vs the guesses. This is for logistic outcome matrices such that the Note that an IllegalArgumentException is thrown if the two passed in matrices aren't the same length.

Parameters:

realOutcomes - the real outcomes (usually binary)

guesses - the guesses (usually a probability vector)

stats

public String stats()

addToConfusion

public void addToConfusion(int real,
                  int guess)

Adds to the confusion matrix

Parameters:

real - the actual guess

guess - the system guess

classCount

public int classCount(int i)

Returns the number of times the given label has actually occurred

Parameters:

i - the label

Returns:

the number of times the label actually occurred

numtimesPredicted

public int numtimesPredicted(int label)

Returns the number of times a given label was predicted

Parameters:

label - the label to get

Returns:

the number of times the given label was predicted

numTimesPredicted

public int numTimesPredicted(int actual,
                    int predicted)

Gets the number of times the given class was predicted for the given predicted label

Parameters:

actual -

predicted -

Returns:

precision

public double precision()

f1

public double f1()

f1

public double f1(int i)

Calculate f1 score for a given class

Parameters:

i - the label to calculate f1 for

Returns:

the f1 score for the given label

recall

public double recall()

Returns the recall for the outcomes

Returns:

the recall for the outcomes

precision

public double precision(int i)

Returns the precision for a given label

Parameters:

i - the label

Returns:

the precision for the label

incrementTruePositives

public void incrementTruePositives()

incrementFalseNegatives

public void incrementFalseNegatives()

incrementFalsePositives

public void incrementFalsePositives(int i)