Artificial Neural Networks

PANN
Estimates the parameters of a tapped-delay, three-layer, feed-forward ANN with polynomial activation functions.

SANN
Estimates the parameters of a tapped-delay, three-layer, feed-forward ANN with sigmoidal activation functions.

PANNS
Estimates the parameters of a PANN architecture with a unipolar sigmoidal function applied to the output.

SPANN
Estimates parameters of a tapped-delay, three-layer, forward ANN with cascaded polynomial-sigmoidal activation functions.

RANN
Computes the output of a PANN, SANN, PANNS or SPANN architecture with given parameter files for a specific input.

KANN
Estimates the 1st and 2nd order Volterra kernels of the nonlinear model defined by PANN, SANN PANNS or SPANN.