error.TAO {AMORE}R Documentation

Neural network training error criteria.

Description

The error functions calculate the goodness of fit of a neural network according to certain criterium:

  • LMS: Least Mean Squares Error.
  • LMLS: Least Mean Log Squares minimization.
  • TAO: TAO error minimization. The deltaE functions calculate the influence functions of their error criteria.

    Usage

    error.LMS(arguments)
    error.LMLS(arguments)
    error.TAO(arguments)
    deltaE.LMS(arguments)
    deltaE.LMLS(arguments)
    deltaE.TAO(arguments)
    

    Arguments

    arguments List of arguments to pass to the functions.
  • The first element is the prediction of the neuron.
  • The second element is the corresponding component of the target vector.
  • The third element is the whole net. This allows the TAO criterium to know the value of the S parameter and eventually ( next minor update) will allow the user to apply regularization criteria.
  • Value

    This functions return the error and influence function criteria.

    Author(s)

    Manuel Castejón Limas. manuel.castejon@unileon.es
    Joaquin Ordieres Meré. joaquin.ordieres@dim.unirioja.es
    Ana González Marcos. ana.gonzalez@unileon.es
    Alpha V. Pernía Espinoza. alpha.pernia@alum.unirioja.es
    Eliseo P. Vergara Gonzalez. eliseo.vergara@dim.unirioja.es
    Francisco Javier Martinez de Pisón. francisco.martinez@dim.unirioja.es
    Fernando Alba Elías. fernando.alba@unavarra.es

    References

    Pernía Espinoza, A.V., Ordieres Meré, J.B., Martínez de Pisón, F.J., González Marcos, A. TAO-robust backpropagation learning algorithm. Neural Networks. Vol. 18, Issue 2, pp. 191–204, 2005.

    Simon Haykin. Neural Networks – a Comprehensive Foundation. Prentice Hall, New Jersey, 2nd edition, 1999. ISBN 0-13-273350-1.

    See Also

    train


    [Package AMORE version 0.2-11 Index]