layer: Describe a layer of a neural network

In davharris/mistnet2: Neural Networks with Latent Random Variables

Description

Describe a layer of a neural network

Usage

layer(activator, n_nodes, weight_prior, weights = NULL, biases = NULL, ...)

Arguments

activator	an activator object, determining the nonlinear function performed by this layer and its derivatives
n_nodes	a whole number determining the size of the layer. The network's last layer should have the same number of nodes as there are columns in y.
weight_prior	a distribution object that acts as a prior distribution for the weight matrix.
weights	an optional weight matrix; if omitted, the mistnet function will randomly initialize the weights
biases	an optional bias vector; if omitted, the mistnet function will randomly initialize the biases
...	Additional objects to include in the layer (not currently used by the mistnet function)

Value

a list of class layer containing the objects above

davharris/mistnet2 documentation built on May 14, 2019, 9:28 p.m.