class TMVA::MethodANNBase: public TMVA::MethodBase

 define the options (their key words) that can be set in the option string
 here the options valid for ALL MVA methods are declared.
 know options: NCycles=xx              :the number of training cycles
               Normalize=kTRUE,kFALSe  :if normalised in put variables should be used
               HiddenLayser="N-1,N-2"  :the specification of the hidden layers
               NeuronType=sigmoid,tanh,radial,linar  : the type of activation function
                                                       used at the neuronn

 decode the options in the option string

 parse layout specification string and return a vector, each entry
 containing the number of neurons to go in each successive layer

 initialize ANNBase object

 destructor

 delete/clear network

 delete a network layer

 build network given a layout (number of neurons in each layer)
 and optional weights array

 build the network layers

 build a single layer with neurons and synapses connecting this
 layer to the previous layer

 add synapses connecting a neuron to its preceding layer

 initialize the synapse weights randomly

 force the synapse weights

 force the input values of the input neurons
 force the value for each input neuron

 calculate input values to each neuron

 print messages, turn off printing by setting verbose and debug flag appropriately

 wait for keyboard input, for debugging

 print network representation, for debugging

 print a single layer, for debugging

 print a neuron, for debugging

 get the mva value generated by the NN

 write the weights stream

 destroy/clear the network then read it back in from the weights file

 compute ranking of input variables by summing function of weights

 write histograms to file

 write specific classifier response

 setters for subclasses

 this will have to be overridden by every subclass

 accessors