Synopsis

A Deep Back-Propagation Neural Network. For more information see: Geoffrey Hinton, Ruslan Salakhutdinov (2006). Reducing the Dimensionality of Data with Neural Networks. Science. 313(5786):504-507.

BibTeX

@article{Hinton2006,
   author = {Geoffrey Hinton and Ruslan Salakhutdinov},
   journal = {Science},
   number = {5786},
   pages = {504-507},
   title = {Reducing the Dimensionality of Data with Neural Networks},
   volume = {313},
   year = {2006}
}

Options

  • -N <value>

    Sets the number of RBMs default: 2

  • -H <value>

    Sets the number of hidden units default: 10

  • -E <value>

    Sets the maximum number of epochs default: 1000 (auto-cut-out)

  • -r <value>

    Sets the learning rate (tyically somewhere between 'very small' and 0.1) default: 0.1

  • -m <value>

    Sets the momentum (typically somewhere between 0.1 and 0.9) default: 0.1

  • -W <classifier name>

    Full name of base classifier. (default: weka.classifiers.trees.J48)

  • -output-debug-info

    If set, classifier is run in debug mode and may output additional info to the console

  • -do-not-check-capabilities

    If set, classifier capabilities are not checked before classifier is built (use with caution).

  • -num-decimal-places

    The number of decimal places for the output of numbers in the model (default 2).

  • -batch-size

    The desired batch size for batch prediction (default 100).

Options specific to classifier weka.classifiers.trees.J48:

  • -U

    Use unpruned tree.

  • -O

    Do not collapse tree.

  • -C <pruning confidence>

    Set confidence threshold for pruning. (default 0.25)

  • -M <minimum number of instances>

    Set minimum number of instances per leaf. (default 2)

  • -R

    Use reduced error pruning.

  • -N <number of folds>

    Set number of folds for reduced error pruning. One fold is used as pruning set. (default 3)

  • -B

    Use binary splits only.

  • -S

    Do not perform subtree raising.

  • -L

    Do not clean up after the tree has been built.

  • -A

    Laplace smoothing for predicted probabilities.

  • -J

    Do not use MDL correction for info gain on numeric attributes.

  • -Q <seed>

    Seed for random data shuffling (default 1).

  • -doNotMakeSplitPointActualValue

    Do not make split point actual value.

  • -output-debug-info

    If set, classifier is run in debug mode and may output additional info to the console

  • -do-not-check-capabilities

    If set, classifier capabilities are not checked before classifier is built (use with caution).

  • -num-decimal-places

    The number of decimal places for the output of numbers in the model (default 2).

  • -batch-size

    The desired batch size for batch prediction (default 100).