Class Loss<TPrediction>

Namespace

NeuralNetworks.Losses

Assembly

NeuralNetworks.dll

The "loss" of a neural network.

public abstract class Loss<TPrediction>

Type Parameters

TPrediction

Inheritance

object

Loss<TPrediction>

Derived

Loss2D

Inherited Members

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

object.ReferenceEquals(object, object)

object.ToString()

Properties

Prediction

public TPrediction Prediction { get; }

Property Value

TPrediction

Target

protected TPrediction Target { get; }

Property Value

TPrediction

Methods

Backward()

Computes gradient of the loss value with respect to the input to the loss function.

public TPrediction Backward()

Returns

TPrediction

CalculateLoss()

protected abstract float CalculateLoss()

Returns

float

CalculateLossGradient()

protected abstract TPrediction CalculateLossGradient()

Returns

TPrediction

Clone()

public Loss<TPrediction> Clone()

Returns

Loss<TPrediction>

CloneBase()

protected virtual Loss<TPrediction> CloneBase()

Returns

Loss<TPrediction>

EnsureSameShape(TPrediction?, TPrediction)

[Conditional("DEBUG")]
protected abstract void EnsureSameShape(TPrediction? prediction, TPrediction target)

Parameters

prediction TPrediction

target TPrediction

Forward(TPrediction, TPrediction)

Computes the actual loss value

public float Forward(TPrediction prediction, TPrediction target)

Parameters

prediction TPrediction

target TPrediction

Returns

float