neural (version 1.4.2.1)

mlp: MLP neural network

Description

The recalling method of the MLP network which was trained by the mlptrain function.

Usage

mlp(inp,weight,dist,neurons,actfns=c(),layer=NaN, ...)

Arguments

inp

a matrix that contains one input data in each row.

weight

the weights of the network.

dist

the distortions of the network.

neurons

a numeric vector with length equals to the number of layers in the network, and the ith layer will contains neurons[i] neuron.

actfns

a list, which contains the numeric code of the activation functions, or the activation function itself. The length of the vector must be the same as the length of the neurons vector, and each element of the vector must be between 1 and 4 or a func

layer

the number of the layer as far as the function is calculating. If the value is NaN, the function will calculate till the output layer. It can be useful if you need the output of one of the hidden layers.

...

currently not used.

Value

• a matrix that contains the response data of the network, each row contains one response.

Details

the "weight, dist, neurons, actfns" arguments can be determined by the mlptrain algorithm.

See Also

`mlptrain' for training an MLP network, `rbf' and `rbftrain' for approximation.