ANN_CONV_W - Function to convert the weight and bias stored in a matrix or vector form in the other form (vector or matrix form respectively)

Calling Sequence

Wtr = ANN_CONV_W(Wini,Nin,Nhid,Nout,Type)

Parameters

• Wini : Initial weight and bias values in vector form (Matrix [Px1] where P = {Nb Input+1}*Nb hidden nodes + {Nb hidden nodes+1}*Nb ouput) or in matrix form(2 dimensions Matrix [max(Nhid,Nout) x max(Nin+1,Nhid+1) x 2]): Wini(1:Nhid,1,1) are the bias for the hidden neurons Wini(1:Nhid,2:P+1,1) are the weights for the hidden neurons (P weights for each hidden neuron) Wini(1:M,1,2) are the bias for the ouput neurons Wini(1:M,2:Nhid+1,2) are the weights for the ouput neurons (Nhid weights for each output neuron)
• Nin : Number of input nodes
• Nhid : Number of hidden nodes
• Nout : Number of output nodes
• Type : Type of Wini ('matrix' or 'vector')
• Wtr : Resulting conversion (if Wini is of 'matrix' type, Wtr is of 'vector' type and vice and versa)

Description

This function offers a convenient way to store network bias and weight values.

The size of the vector form gives immediately the number of parameters required by a network.

Examples

   // Ouput from a network with 6 input nodes, 4 nodes in the hidden layer and 1 output node
   IN = rand(6,100);
   W  = rand(4,7,2);
   [OUT,IN_W,HID_OUT] = ANN_SIM(IN,4,1,W);
   Wvect  = ANN_CONV_W(W,6,4,1,'matrix');

  

See Also

ANN_LMBR ,   ANN_JACOB ,   ANN_NORM ,   ANN_SIM ,  

Authors

Julien Lerat CEMAGREF Antony, HBAN Unit, julien.lerat@cemagref.fr