Neural Network Toolbox™

Product net input function

Syntax

• N = netprod({Z1,Z2,...,Zn},FP)
  dN_dZj = netprod('dz'j,Z,N,FP)
  info = netprod(code)
  

Description

netprod is a net input function. Net input functions calculate a layer's net input by combining its weighted inputs and biases.

netprod(Z1,Z2,...,Zn) takes

Zi	S x Q matrices in a row cell array
FP	Row cell array of function parameters (optional, ignored)

and returns an elementwise product of Z1 to Zn.

netprod(code) returns information about this function. The following codes are defined:

'deriv'	Name of derivative function
'fullderiv'	Full N x S x Q derivative = 1, elementwise S x Q derivative = 0
'name'	Full name
'fpnames'	Returns names of function parameters
'fpdefaults'	Returns default function parameters

Examples

Here netprod combines two sets of weighted input vectors (user-defined).

• z1 = [1 2 4;3 4 1];
  z2 = [-1 2 2; -5 -6 1];
  z = {z1,z2};
  n = netprod({Z})
  

Here netprod combines the same weighted inputs with a bias vector. Because Z1 and Z2 each contain three concurrent vectors, three concurrent copies of B must be created with concur so that all sizes match.

• b = [0; -1];
  z = {z1, z2, concur(b,3)};
  n = netprod(z)
  

Network Use

You can create a standard network that uses netprod by calling newpnn or newgrnn.

To change a network so that a layer uses netprod, set net.layers{i}.netInputFcn to 'netprod'.

In either case, call sim to simulate the network with netprod. See newpnn or newgrnn for simulation examples.

See Also

network/sim, netsum, concur

Provide feedback about this page

netinv

netsum

© 1984-2008- The MathWorks, Inc.    -   Site Help   -   Patents   -   Trademarks   -   Privacy Policy   -   Preventing Piracy   -   RSS