Rbm

Value Members

1. final def !=(arg0: AnyRef): Boolean

   Definition Classes

   AnyRef

2. final def !=(arg0: Any): Boolean

   Definition Classes

   Any

3. final def ##(): Int

   Definition Classes

   AnyRef → Any

4. final def ==(arg0: AnyRef): Boolean

   Definition Classes

   AnyRef

5. final def ==(arg0: Any): Boolean

   Definition Classes

   Any

6. def activities(input: Mat): List[Mat]

   Definition Classes

   NeuralNetLike

7. def andThen[A](g: (Mat) ⇒ A): (Mat) ⇒ A

   Definition Classes

   Function1

   Annotations

   @unspecialized()

8. def apply(input: Mat): Mat

   Propagates the input from the visible layer up to the top layer

   Propagates the input from the visible layer up to the top layer

   Definition Classes

   NeuralNetLike → Function1

9. final def asInstanceOf[T0]: T0

   Definition Classes

   Any

10. def build(ls: List[Layer]): Rbm

    Builds a neural net of the right type and of the right shape out of specified layers.

    Builds a neural net of the right type and of the right shape out of specified layers.

    Note that this method depends on instance, not just a class: fore example the Autoencoder has to know what it's 'central' Layer is.

    Definition Classes

    Rbm → NeuralNet → NeuralNetLike

11. def clone(): Rbm

    Definition Classes

    Rbm → AnyRef

12. def compose[A](g: (A) ⇒ Mat): (A) ⇒ Mat

    Definition Classes

    Function1

    Annotations

    @unspecialized()

13. def confabulate(hiddenActivation: Mat, sampleVisibleUnitsDeterministically: Boolean = false): Mat

    Given hidden activation, samples a hidden state, calculates visible activation, and samples visible state.

    Given hidden activation, samples a hidden state, calculates visible activation, and samples visible state.

    If sampleVisibleStatesDeterministically is set to true, activations of visible neurons are returned directly, no random sampling occurs for visible neurons in this case.

14. val connection: FullBipartiteConnection

15. def contrastiveDivergence(minibatch: Mat, steps: Int = 1, sampleVisibleUnitsDeterministically: Boolean = false): (Mat, Mat, Mat)

    Calculates (very coarse) approximations of partial derivatives of the logarithmized probability of the minibatch w.

    Calculates (very coarse) approximations of partial derivatives of the logarithmized probability of the minibatch w.r.t. biases of the unit layers and weights of the connection layer.

    returns

    tuple containing three matrices that correspond to derivatives w.r.t. biases of visible units, weights, and hidden units respectively.

    Attributes

    protected

16. var dataSample: Option[Mat]

    Definition Classes

    NeuralNetLike

17. final def eq(arg0: AnyRef): Boolean

    Definition Classes

    AnyRef

18. def equals(arg0: Any): Boolean

    Definition Classes

    AnyRef → Any

19. def extractStatistics(visibleStates: Mat, hiddenActivations: Mat): (Mat, Mat, Mat)

    Extracts average visible unit state, average hidden unit activation, and average products of visible states and hidden activations from the samples of visible units and hidden activations.

    Extracts average visible unit state, average hidden unit activation, and average products of visible states and hidden activations from the samples of visible units and hidden activations.

    The dimensions of extracted matrices correspond to dimensions of visible biases, weight Mat, and hidden biases respectively.

    Attributes

    protected

20. def finalize(): Unit

    Attributes

    protected[java.lang]

    Definition Classes

    AnyRef

    Annotations

    @throws( classOf[java.lang.Throwable] )

21. final def getClass(): Class[_]

    Definition Classes

    AnyRef → Any

22. def gibbsSampling(visibleStates: Mat, steps: Int = 1, sampleVisibleUnitsDeterministically: Boolean = false): (Mat, Mat)

    Performs Gibbs sampling steps times, starting with the input clamped to the visible layer.

    Performs Gibbs sampling steps times, starting with the input clamped to the visible layer.

    If steps is zero, the visible input together with hidden layer activation is returned (useful for collecting positive statistics in contrastive divergence).

    If steps is greater than zero, returns visible reconstruction and exact hidden layer activations after specified number of steps.

    If sampleVisibleUnitsDeterministically is set to true, then activations of the visible units are used instead of random samples.

    The hidden units are always updated randomly.

23. def hashCode(): Int

    Definition Classes

    AnyRef → Any

24. val hidden: RbmLayer

25. final def isInstanceOf[T0]: Boolean

    Definition Classes

    Any

26. val layers: List[Layer]

    Enumerates layers of this (linear) neural net.

    Enumerates layers of this (linear) neural net.

    TODO: generalize it to arbitrary directed acyclic graphs, what's so special about lists?...

    Definition Classes

    NeuralNet → NeuralNetLike

27. def layersToString(layers: List[Layer]): String

    Attributes

    protected

    Definition Classes

    NeuralNet

28. final def ne(arg0: AnyRef): Boolean

    Definition Classes

    AnyRef

29. final def notify(): Unit

    Definition Classes

    AnyRef

30. final def notifyAll(): Unit

    Definition Classes

    AnyRef

31. def optimize(input: Mat, output: Mat, errorFunctionFactory: DifferentiableErrorFunctionFactory[Mat] = SquareErrorFunctionFactory, relativeValidationSetSize: Double, maxEvals: Int, trainingObservers: List[TrainingObserver]): Rbm

    Performs optimization of all parameters of the neural network using the specified input and output, the specified method to define an error function (defaults to SquareErrorFunctionFactory).

    Performs optimization of all parameters of the neural network using the specified input and output, the specified method to define an error function (defaults to SquareErrorFunctionFactory).

    Standard feed-forward algorithm is used for evaluation of the function, backpropagation is used for calculation of the gradient.

    Definition Classes

    NeuralNetLike

32. def prependAffineLinearTransformation(factor: Mat, offset: Mat): Rbm

    Assumes that this is a "usual" neural net with alternating unit and connection layers and prepends an affine linear transformation to it.

    Assumes that this is a "usual" neural net with alternating unit and connection layers and prepends an affine linear transformation to it.

    Why the heck did I implement biased layers at all, why didn't I stuff all this cruft into something like "AffineLinearTransform" or so... Damn

    Definition Classes

    NeuralNetLike

33. def reinitialize(config: RbmTrainingConfiguration): Rbm

34. def reverse(output: Mat): Mat

    Propagates the output from top layer down to the visible layer

    Propagates the output from top layer down to the visible layer

    Definition Classes

    NeuralNetLike

35. final def synchronized[T0](arg0: ⇒ T0): T0

    Definition Classes

    AnyRef

36. def toImage: BufferedImage

    Definition Classes

    NeuralNetLike → Visualizable

37. def toImage(w: Int, h: Int): BufferedImage

    Definition Classes

    Visualizable

38. def toImage(colormap: (Double) ⇒ Int): BufferedImage

    Definition Classes

    Visualizable

39. def toString(): String

    Definition Classes

    Rbm → NeuralNet → Function1 → AnyRef → Any

40. def train[Fitness](trainingSet: Mat, configuration: RbmTrainingConfiguration, trainingObservers: List[TrainingObserver] = Nil, terminationCriterion: TerminationCriterion[Rbm, Int], resultSelector: ResultSelector[Rbm, Fitness])(implicit arg0: Ordering[Fitness]): Rbm

    Trains this Rbm with the data contained in the minibatches with parameters specified in the configuration.

    Trains this Rbm with the data contained in the minibatches with parameters specified in the configuration.

    Returns the training data processed by the trained Rbm

41. val visible: RbmLayer

42. final def wait(): Unit

    Definition Classes

    AnyRef

    Annotations

    @throws( ... )

43. final def wait(arg0: Long, arg1: Int): Unit

    Definition Classes

    AnyRef

    Annotations

    @throws( ... )

44. final def wait(arg0: Long): Unit

    Definition Classes

    AnyRef

    Annotations

    @throws( ... )