eVision.EasyDeepLearning.EROCPoint Struct Reference

The structure representing a point on the ROC (Receiver Operating Characteristic) curve. More...

Public Member Functions
	EROCPoint (float threshold, int truePositiveCount, int positiveCount, int falsePositiveCount, int negativeCount)
	Constructs a EROCPoint.
void	Load (string path)
	Loads the ROC point. The given ESerializer must have been created for reading.
void	Load (eVision.ESerializer serializer)
	Loads the ROC point. The given ESerializer must have been created for reading.
void	Save (string path)
	Saves the ROC point. The given ESerializer must have been created for writing.
void	Save (eVision.ESerializer serializer)
	Saves the ROC point. The given ESerializer must have been created for writing.

Public Attributes
float	TPR
	The True Positive Rate: it is the number of defective images classified as defective, normalized by the number of defective images.
int	TP
	The True Positive count: it is the number of defective images classified as defective.
int	P
	The Positive count: it is the number of defective images.
float	FPR
	The False Positive Rate: it is the number of good image classified as defective, normalized by the number of good images.
int	FP
	The False Positive count: it is the number of good image classified as defective.
int	N
	The Negative count: it is the number of good images.
float	Threshold
	The threshold associated with the true and false positive rates (see EUnsupervisedSegmenter.ClassificationThreshold).

Detailed Description

The structure representing a point on the ROC (Receiver Operating Characteristic) curve.

License(s): None

Constructor & Destructor Documentation

EROCPoint()

eVision.EasyDeepLearning.EROCPoint.EROCPoint	(	float	threshold,
		int	truePositiveCount,
		int	positiveCount,
		int	falsePositiveCount,
		int	negativeCount )

Constructs a EROCPoint.

Parameters

threshold	Threshold corresponding to the ROC point.
truePositiveCount	Number of defective images classified as defective.
positiveCount	Total number of defective images.
falsePositiveCount	Number of good images classified as defective.
negativeCount	Total number of good images.

Member Function Documentation

Load() [1/2]

void eVision.EasyDeepLearning.EROCPoint.Load

(

eVision.ESerializer

serializer

)

Loads the ROC point. The given ESerializer must have been created for reading.

Parameters

serializer

The serializer.

Load() [2/2]

void eVision.EasyDeepLearning.EROCPoint.Load

(

string

path

)

Loads the ROC point. The given ESerializer must have been created for reading.

Parameters

path	The file path.

Save() [1/2]

void eVision.EasyDeepLearning.EROCPoint.Save

(

eVision.ESerializer

serializer

)

Saves the ROC point. The given ESerializer must have been created for writing.

Parameters

serializer

The serializer.

Save() [2/2]

void eVision.EasyDeepLearning.EROCPoint.Save

(

string

path

)

Saves the ROC point. The given ESerializer must have been created for writing.

Parameters

path	The file path.

Member Data Documentation

FP

int eVision.EasyDeepLearning.EROCPoint.FP

The False Positive count: it is the number of good image classified as defective.

FPR

float eVision.EasyDeepLearning.EROCPoint.FPR

The False Positive Rate: it is the number of good image classified as defective, normalized by the number of good images.

N

int eVision.EasyDeepLearning.EROCPoint.N

The Negative count: it is the number of good images.

P

int eVision.EasyDeepLearning.EROCPoint.P

The Positive count: it is the number of defective images.

Threshold

float eVision.EasyDeepLearning.EROCPoint.Threshold

The threshold associated with the true and false positive rates (see EUnsupervisedSegmenter.ClassificationThreshold).

TP

int eVision.EasyDeepLearning.EROCPoint.TP

The True Positive count: it is the number of defective images classified as defective.

TPR

float eVision.EasyDeepLearning.EROCPoint.TPR

The True Positive Rate: it is the number of defective images classified as defective, normalized by the number of defective images.

The TPR can be NaN (Not A Number) if the metrics were computed using only good images.