PatientLevelPrediction (version 4.3.10)

setMLP: Create setting for neural network model with python

Description

Create setting for neural network model with python

Usage

setMLP(
  size = 4,
  alpha = c(0.3, 0.01, 0.001, 1e-06),
  maxIter = 2000,
  tol = 1e-04,
  learningRateInit = 0.001,
  nIterNoChange = 10,
  beta1 = 0.9,
  beta2 = 0.999,
  epsilon = c(1, 0.1, 1e-08),
  seed = NULL
)

Arguments

size

The number of hidden nodes

alpha

The l2 regularisation

maxIter

Maximum number of iterations. The solver iterates until convergence (determined by <U+2018>tol<U+2019>) or this number of iterations.

tol

Tolerance for the optimization

learningRateInit

The initial learning rate used. It controls the step-size in updating the weights.

nIterNoChange

Maximum number of epochs to not meet tol improvement.

beta1

Exponential decay rate for estimates of first moment vector in adam, should be in [0, 1).

beta2

Exponential decay rate for estimates of second moment vector in adam, should be in [0, 1).

epsilon

Value for numerical stability in adam.

seed

A seed for the model

Examples

# NOT RUN {
model.mlp <- setMLP(size=4, alpha=0.00001, seed=NULL)
# }