Get weights for the skip layer in a neural network, only valid for networks created using skip = TRUE with the nnet function.
neuralskips(mod_in, ...)
# S3 method for nnet
neuralskips(mod_in, rel_rsc = NULL, ...)input object for which an organized model list is desired.
arguments passed to other methods
numeric indicating the scaling range for the width of connection weights in a neural interpretation diagram. Default is NULL for no rescaling. Scaling is relative to all weights, not just those in the primary network.
Returns a list of connections for each output node, where each element of the list is the connection for each input node in sequential order to the respective output node. The first weight in each element is not the bias connection, unlike the results for neuralweights.
This function is similar to neuralweights except only the skip layer weights are returned.
data(neuraldat)
set.seed(123)
## using nnet
library(nnet)
mod <- nnet(Y1 ~ X1 + X2 + X3, data = neuraldat, size = 5, linout = TRUE,
skip = TRUE)
#> # weights: 29
#> initial value 2013.863122
#> iter 10 value 0.826394
#> iter 20 value 0.014326
#> iter 30 value 0.000585
#> final value 0.000082
#> converged
neuralskips(mod)
#> $`out 1`
#> [1] -0.07848376 -0.10607251 0.05082103
#>