layer_spline: Wrapper function to create spline layer

In davidruegamer/deepregression: Fitting Deep Distributional Regression

Wrapper function to create spline layer

Description

Wrapper function to create spline layer

Usage

layer_spline(
  object,
  name = NULL,
  trainable = TRUE,
  input_shape,
  regul = NULL,
  Ps,
  use_bias = FALSE,
  kernel_initializer = "glorot_uniform",
  bias_initializer = "zeros",
  variational = FALSE,
  posterior_fun = NULL,
  diffuse_scale = 1000,
  output_dim = 1L,
  k_summary = k_sum,
  ...
)

Arguments

object	or layer object.
name	An optional name string for the layer (should be unique).
trainable	logical, whether the layer is trainable or not.
input_shape	Input dimensionality not including samples axis.
regul	if set to 0, no regularization is applied
Ps	list of penalty matrices times lambdas
use_bias	whether or not to use a bias in the layer. Default is FALSE.
kernel_initializer	function to initialize the kernel (weight). Default is "glorot_uniform".
bias_initializer	function to initialize the bias. Default is 0.
variational	logical, if TRUE, priors corresponding to the penalties and posteriors as defined in posterior_fun are created
posterior_fun	function defining the variational posterior
diffuse_scale	diffuse scale prior for scalar weights
output_dim	the number of units for the layer
k_summary	keras function for the penalty (see ?deepregression for details)
...	further arguments passed to args used in create_layer

davidruegamer/deepregression documentation built on May 30, 2022, 6:21 p.m.