ANM_MLP: Additive Noise Model using MLP regression.

In DMIRLAB-Group/CANM: Causal Discovery with Cascade Nonlinear Additive Noise Models

Description

Additive noise model using the MLP regression.

Usage

ANM_MLP(x, y, epochs = 20L, D_in = 1L, D_H1 = 7L, D_H2 = 5L,
  D_out = 1L, batch_size = 128L, cuda = FALSE, seed = 0L,
  log_interval = 10L, learning_rate = 0.01, verbose = FALSE, ...)

Arguments

x	The observation of the cause.
y	The observation of the effect.
epochs	The training epochs
D_in	Dimension of input
D_H1	Dimension of hidden layer 1
D_H2	Dimension of hidden layer 2
D_out	Dimension of output
batch_size	Batch size
cuda	Use GPU
seed	Random Seed
log_interval	The option of verbose that output the training detail at each interval.
learning_rate	The learning rate.
verbose	print the output
...	Other parameters

Examples

set.seed(0)
x=rnorm(1000)
y=exp(x)+rnorm(1000)
result=ANM_MLP(x,y)
if(result$HSIC_xy<result$HSIC_yx){
 print("X->Y")
}else{
 print("Y->X")
}

DMIRLAB-Group/CANM documentation built on May 31, 2019, 12:40 a.m.