initcon :: Functions (Neural Network Toolbox™)

initcon

Purpose

Conscience bias initialization function

initcon is a bias initialization function that initializes biases for learning with the learncon learning function.

initcon (S,PR) takes two arguments,

S
Number of rows (neurons)

PR
R x 2 matrix of R = [Pmin Pmax] (default = [1 1])

`S`	Number of rows (neurons)
`PR`	`R` x `2` matrix of `R = [Pmin Pmax]` (default = `[1 1]`)

and returns an S x 1 bias vector.

Note that for biases, R is always 1. initcon could also be used to initialize weights, but it is not recommended for that purpose.

Here initial bias values are calculated for a five-neuron layer.

You can create a standard network that uses initcon to initialize weights by calling newc.

To prepare the bias of layer i of a custom network to initialize with initcon,

Set net.initFcn to 'initlay'. (net.initParam automatically becomes initlay's default parameters.)
Set net.layers{i}.initFcn to 'initwb'.
Set net.biases{i}.initFcn to 'initcon'.

To initialize the network, call init. See newc for initialization examples.

learncon updates biases so that each bias value b(i) is a function of the average output c(i) of the neuron i associated with the bias.

initcon gets initial bias values by assuming that each neuron has responded to equal numbers of vectors in the past.