MultiClassification.Probabilities Method

MultiClassificationProbabilities Method

Returns classification probabilities for the input pattern x.

Namespace: Imsl.DataMining.Neural
Assembly: ImslCS (in ImslCS.dll) Version: 6.5.2.0

Syntax

public virtual double[] Probabilities(
	double[] x
)

Public Overridable Function Probabilities ( 
	x As Double()
) As Double()

public:
virtual array<double>^ Probabilities(
	array<double>^ x
)

abstract Probabilities : 
        x : float[] -> float[] 
override Probabilities : 
        x : float[] -> float[]

Parameters

x: Type: SystemDouble
A double array containing the input patterns to classify. The length of x must be equal to the number of input nodes.

Return Value

Type: Double
A double containing the scaled probabilities.

Remarks

The number of probabilities is equal to the number of target classes, which is the number of outputs in the FeedForwardNetwork. Each are calculated using the softmax activation for each of the OutputPerceptrons. The softmax function transforms the outputs potential z to the probability y by

$y_i={{\rm{softmax}}_{\rm{i}}=\frac{{{\mathop{\rm e}\nolimits}^{Z_i}}} {{\sum\limits_{j=1}^C{e^{Z_j}}}}}$

Reference

MultiClassification Class

Imsl.DataMining.Neural Namespace